Change search
Refine search result
9899100101102103 5001 - 5050 of 5121
CiteExportLink to result list
Permanent link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Rows per page
  • 5
  • 10
  • 20
  • 50
  • 100
  • 250
Sort
  • Standard (Relevance)
  • Author A-Ö
  • Author Ö-A
  • Title A-Ö
  • Title Ö-A
  • Publication type A-Ö
  • Publication type Ö-A
  • Issued (Oldest first)
  • Issued (Newest first)
  • Created (Oldest first)
  • Created (Newest first)
  • Last updated (Oldest first)
  • Last updated (Newest first)
  • Standard (Relevance)
  • Author A-Ö
  • Author Ö-A
  • Title A-Ö
  • Title Ö-A
  • Publication type A-Ö
  • Publication type Ö-A
  • Issued (Oldest first)
  • Issued (Newest first)
  • Created (Oldest first)
  • Created (Newest first)
  • Last updated (Oldest first)
  • Last updated (Newest first)
Select
The maximal number of hits you can export is 250. When you want to export more records please use the 'Create feeds' function.
  • 5001.
    Zeynep, Ahmet
    et al.
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Väänänen-Vainio Mattila, Kaisa
    Tampere University of Technology, , .
    Face to face makes a difference: Recommendation Practices of Users of Mobile Services2011In: UbiComp2011, Beijing, China: ACM Press , 2011Conference paper (Other academic)
    Abstract [en]

    The mobile app stores and markets provide companies, independent developers and researchers alike with possibilities to distribute innovative designs for mobile devices on a global scale. However, reaching a large numbers of users does not in itself ensure a large number of users adopting the mobile application or service. Large-scale adoption depends on additional factors such as novelty in service design, ease of use, enjoyable interaction, built-in mechanisms for further distribution of the mobile service as well as the practice of word-of-mouth recommendations. In this position paper we present the background and preliminary findings from a study aimed at investigating the motivations and practices by which users recommend mobile apps and services among their acquaintances. We discuss our perspective on distribution of mobile applications and services on a large scale and end this paper by suggesting questions for discussion and future research.

  • 5002.
    Zhang, Ge
    et al.
    Fraunhofer Institute FOKUS, Berlin, Germany.
    Ehlert, Sven
    Fraunhofer Institute FOKUS, Berlin, Germany.
    Magedanz, Thomas
    Fraunhofer Institute FOKUS, Berlin, Germany.
    Sisalem, Dorgham
    Tekelec, Berlin, Germany.
    Denial of Service Attack and Prevention on SIP VoIP Infrastructures Using DNS Flooding2007In: Proceedings of the 1st International Conference on Principles, Systems and Applications of IP Telecommunications (IPTCOMM 2007), New York: ACM Press, 2007, 57-66 p.Conference paper (Refereed)
    Abstract [en]

    A simple yet effective Denial of Service (DoS) attack on SIP servers is to flood the server with requests addressed at irresolvable domain names.

  • 5003.
    Zhang, Ge
    et al.
    Karlstad University, Faculty of Economic Sciences, Communication and IT, Department of Computer Science.
    Fischer-Hübner, Simone
    Karlstad University, Faculty of Economic Sciences, Communication and IT, Department of Computer Science.
    Martucci, Leonardo A.
    Karlstad University, Faculty of Economic Sciences, Communication and IT, Department of Computer Science.
    Ehlert, Sven
    Fraunhofer FOKUS, Berlin, Germany.
    Revealing the calling history on SIP VoIP systems by timing attacks2009In: Proceedings of the 4th International Conference on Availability, Reliability and Security (ARES 2009), IEEE Press, IEEE Computer Society, 2009, 135-142 p.Conference paper (Refereed)
    Abstract [en]

    Many emergent security threats which did not exist in the traditional telephony network are introduced in SIP VoIP services. To provide high-level security assurance to SIP VoIP services, an inter-domain authentication mechanism is defined in RFC 4474. However, this mechanism introduces another vulnerability: a timing attack which can be used for effectively revealing the calling history of a group of VoIP users. The idea here is to exploit the certificate cache mechanisms supported by SIP VoIP infrastructures, in which the certificate from a caller's domain will be cached by the callee's proxy to accelerate subsequent requests. Therefore, SIP processing time varies depending whether the two domains had been into contact beforehand or not. The attacker can thus profile the calling history of a SIP domain by sending probing requests and observing the time required for processing. The result of our experiments demonstrates that this attack can be easily launched. We also discuss countermeasures to prevent such attacks

  • 5004. Zhang, Pengyi
    et al.
    Liu, Chang
    Hansen, Preben
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    I Need More Time!: The Influence of Native Language on Search Behavior and Experience2016In: Working Notes of CLEF 2016 - Conference and Labs of the Evaluation forum: CEUR Proceedings / [ed] Krisztian Balog, Linda Cappellato, Nicola Ferro, Craig Macdonald, 2016, 1166-1182 p.Conference paper (Refereed)
  • 5005.
    Zhang, Sheng
    et al.
    Nanchang Hangkong Univ, Sch Informat Engn, Nanchang, Peoples R China..
    Wang, Xin
    Nanchang Hangkong Univ, Sch Informat Engn, Nanchang, Peoples R China..
    Yao, Minghui
    Nanchang Hangkong Univ, Sch Informat Engn, Nanchang, Peoples R China..
    Song, William Wei
    Dalarna University, School of Technology and Business Studies, Information Systems. Dalarna Univ, Business Intelligence & Informat, Borlange, Sweden..
    Community-based message transmission with energy efficient in opportunistic networks2016In: Web Information Systems Engineering – WISE 2016: 17th International Conference, Shanghai, China, November 8-10, 2016, Proceedings, Part II, 2016, Vol. 10042, 411-423 p.Conference paper (Refereed)
    Abstract [en]

    An Opportunistic Networks is a wireless self-organized network, in which there is no need to build a fixed connectivity between source node and destination node, and the communication depends on the opportunity of node meeting. There are some classical message transmission algorithms, such as PRoPHET, MaxProp, and so on. In the Opportunity Networks with community characteristic, the different message transmission strategies can be sued in inter-community and intra-community. It improves the message successful delivery ratio significantly. The classical algorithms are CMTS and CMOT. We propose an energy efficient message forwarding algorithm (EEMF) for community-based Opportunistic Networks in this paper. When a message is transmitted, we consider not only the community characteristic, but also the residual energy of each node. The simulation results show that the EEMF algorithm can improve the message successful delivery ratio and reduce the network overhead obviously, in comparison with classical routing algorithms, such as PRoPHET, MaxProp, CMTS and CMOT. Meanwhile the EEMF algorithm can reduce the node's energy consumption and prolong the lifetime of network.

  • 5006.
    Zhang, Yanqing
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Constructing Swedish Fashion Identity Exhibition Review: Svenskt Mode (Swedish Fashion): 2000–2015, Sven-Harry’s Art Museum, Stockholm, May 23–August 31, 20142015In: Fashion theory, ISSN 1362-704X, E-ISSN 1751-7419Article in journal (Refereed)
    Abstract [en]

    This is an exhibition review article. It describes the goal, the curatorship, display and exhibited objects. The exhibition focuses on Swedish fashion design and treats fashion as something in relation to art. It analyzes the exhibition as part of institutionalization of Swedish fashion system. Understanding fashion as a system rather than material clothing is the key point here.

  • 5007.
    Zhang, Yanqing
    et al.
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Juhlin, Oskar
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Fashion as System or Action Net in ‘Fashion in All Things’: a Case in Color Design of Mobile Phones2012In: Fashion: Exploring Critical Issues / [ed] Barbara Brownie, Laura Petican, Johannes Reponen, Oxford: Inter-Disciplinary Press, 2012, no 1, 263-270 p.Chapter in book (Other academic)
    Abstract [en]

    Contemporary fashion has permeated into all things in life beyond clothes. Recently, fashion theories take on interests in organization and system. Kawamura proposes a fashion system through which clothing is transformed into the idea of fashion. Can this fashion system be used to analyze other things in fashion? We present a study using mobile phone, one of the most intimate gadgets to people, as a way to approach ‘fashion in all things’. We chose the color as a way to study the fashion aspect in mobile design. Through the empirical study, we find that the decision making of color in mobile industry is a collective process. It is greatly influenced by technology, materials, consumer lifestyle and trend. The trendy colors in mobile design are not defined by certain cultural or social institutions, but formulated by actions conducted by various actors in certain social context. Our study shows that fashion can embrace more than Kawamura’s system, e.g. the action net of color design in mobile technology. Although mobile design shares some similarities with clothing fashion, the concept of fashion-ology is very Parisian and deals with only clothing. It is not fully applicable to mobile industry. If we want to use a fashion system that can apply to fashion in all things, we should revisit the theory to reveal the general characteristics of the fashion world or build smaller theory for each category.

  • 5008.
    Zhang, Yanqing
    et al.
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Juhlin, Oskar
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    The “life and death” of great Finnish fashion phone: A periodization of changing styles in Nokia phone design between 1992 and 20132016In: Mobile Media & Communication, ISSN 2050-1579, Vol. 4, no 3, 385-404 p.Article in journal (Refereed)
    Abstract [en]

    Visual aesthetics is an essential part of our experience of mobile devices, but the ways in which it is accounted for in design have largely been overlooked. We investigate whether an aesthetization of mobile design is taking place and, if so, how it is being pursued through institutional practices in organizations. We conduct a visual analysis of all Nokia phone releases between 1992 and 2013 complemented by an interview series with key actors. The study reveals a continuous increase in aesthetic variation between 1998 and 2008, which is visible in the variation of colors, forms and materials. The period between 2003 and 2008, which we term the Grand period, marks the peak of aesthetization of Nokia’s devices. It exhibits great variation, and is visibly similar to aesthetics in the fashion industry. With the introduction of the slate form, we see a decrease in visual variation between 2009 and 2013. The interviews reveal how the visual design was driven by organizational strategies, such as customer segmentation in general, and an orientation toward the fashion industry, e.g. in the creation of a fashion segment. The study reveals how aesthetic variation is weaved into a complex innovation system with sometimes conflicting demands deriving from e.g. technology and user interaction.

  • 5009.
    Zhang, Zhi
    KTH, School of Information and Communication Technology (ICT), Electronic, Computer and Software Systems, ECS.
    Hierarchical multi-reader RFID systems for Internet-of-Things2010Licentiate thesis, comprehensive summary (Other academic)
  • 5010.
    Zhang, Zhi
    et al.
    KTH, School of Information and Communication Technology (ICT), Electronic Systems.
    Lu, Zhonghai
    KTH, School of Information and Communication Technology (ICT), Electronic Systems.
    Chen, Qiang
    KTH, School of Information and Communication Technology (ICT), Electronic, Computer and Software Systems, ECS.
    Yan, Xiaolang
    Institute of VLSI Design, Zhejiang University, Hangzhou, China.
    Zheng, Li-Rong
    A high performance multi-reader passive RFID system for Internet-of-ThingsIn: IEEE Transactions on Wireless Communications, ISSN 1536-1276, E-ISSN 1558-2248Article in journal (Other academic)
  • 5011.
    Zhang, Zhi
    et al.
    KTH, School of Information and Communication Technology (ICT), Centres, VinnExcellence Center for Intelligence in Paper and Packaging, iPACK. KTH, School of Information and Communication Technology (ICT), Electronic Systems.
    Lu, Zhonghai
    KTH, School of Information and Communication Technology (ICT), Centres, VinnExcellence Center for Intelligence in Paper and Packaging, iPACK. KTH, School of Information and Communication Technology (ICT), Electronic Systems.
    Chen, Qiang
    KTH, School of Information and Communication Technology (ICT), Centres, VinnExcellence Center for Intelligence in Paper and Packaging, iPACK. KTH, School of Information and Communication Technology (ICT), Electronic Systems.
    Yan, Xiaolang
    Institute of VLSI Design, Zhejiang University, Hangzhou, China.
    Zheng, Li-Rong
    KTH, School of Information and Communication Technology (ICT), Centres, VinnExcellence Center for Intelligence in Paper and Packaging, iPACK. KTH, School of Information and Communication Technology (ICT), Electronic Systems.
    COSMO: CO-simulation with MATLAB and OMNeT++ for indoor wireless networks2010In: 2010 IEEE GLOBAL TELECOMMUNICATIONS CONFERENCE GLOBECOM 2010, 2010Conference paper (Refereed)
    Abstract [en]

    Simulations are widely used to design and evaluate new protocols and applications of indoor wireless networks. However, the available network simulation tools face the challenges of providing accurate indoor channel models, three-dimensional (3-D) models, model portability, and effective validation. In order to overcome these challenges, this paper presents a new CO-Simulation framework based on MATLAB and OMNeT++ (COSMO) to rapidly build credible simulations for indoor wireless networks. A hierarchical ad hoc passive RFID network for indoor tag locating is described as a case study, demonstrating the significance and efficiency of COSMO compared with other network simulators. COSMO surpasses other network simulators in terms of workload and validity.

  • 5012.
    Zhang, Zhi
    et al.
    KTH, School of Information and Communication Technology (ICT), Electronic, Computer and Software Systems, ECS.
    Pang, Zhibo
    KTH, School of Information and Communication Technology (ICT), Electronic, Computer and Software Systems, ECS.
    Chen, Jun
    KTH, School of Information and Communication Technology (ICT), Electronic, Computer and Software Systems, ECS.
    Chen, Qiang
    KTH, School of Information and Communication Technology (ICT), Electronic, Computer and Software Systems, ECS.
    Tenhunen, Hannu
    KTH, School of Information and Communication Technology (ICT), Electronic, Computer and Software Systems, ECS.
    Zheng, Li-Rong
    KTH, School of Information and Communication Technology (ICT), Centres, VinnExcellence Center for Intelligence in Paper and Packaging, iPACK. KTH, School of Information and Communication Technology (ICT), Electronic, Computer and Software Systems, ECS.
    Yan, Xiaolang
    Institute of VLSI Design, Zhejiang University, Hangzhou, China.
    Two-layered wireless sensor networks for warehouses and supermarkets2009Conference paper (Other academic)
    Abstract [en]

    The rapid development of wireless sensor network and RFID technologies offers a wide range of novel applications and services. In this paper, we present a two-layered wireless network for warehouses and supermarkets to monitor goods storage and sale, and assist for quality management and market analysis. The hierarchical architecture uses IEEE 802.15.4a impulse ultra-wideband radio (IR-UWB) communication protocol between slave sensor nodes and master sensor nodes, and IEEE 802.11b/g between master sensor nodes and server. The performance of our proposal is evaluated based on the widely used OMNeT++ simulation environment. Simulation results are presented and discussed according to different sampling rates and traffic loads for specific scenarios requirements. © 2009 IEEE.

  • 5013.
    Zhang, Ziqi
    et al.
    University of Sheffield, England.
    Gentile, Anna Lisa
    University of Sheffield, England.
    Blomqvist, Eva
    Linköping University, Department of Computer and Information Science, Human-Centered systems. Linköping University, Faculty of Science & Engineering.
    Augenstein, Isabelle
    University of Sheffield, England.
    Ciravegna, Fabio
    University of Sheffield, England.
    An Unsupervised Data-driven Method to Discover Equivalent Relations in Large Linked Datasets2017In: Semantic Web, ISSN 1570-0844, E-ISSN 2210-4968, Vol. 8, no 2Article in journal (Refereed)
    Abstract [en]

    This article addresses a number of limitations of state-of-the-art methods of Ontology Alignment: 1) they primarily address concepts and entities while relations are less well-studied; 2) many build on the assumption of the well-formedness of ontologies which is unnecessarily true in the domain of Linked Open Data; 3) few have looked at schema heterogeneity from a single source, which is also a common issue particularly in very large Linked Dataset created automatically from heterogeneous resources, or integrated from multiple datasets. We propose a domain-and language-independent and completely unsupervised method to align equivalent relations across schemata based on their shared instances. We introduce a novel similarity measure able to cope with unbalanced population of schema elements, an unsupervised technique to automatically decide similarity threshold to assert equivalence for a pair of relations, and an unsupervised clustering process to discover groups of equivalent relations across different schemata. Although the method is designed for aligning relations within a single dataset, it can also be adapted for cross-dataset alignment where sameAs links between datasets have been established. Using three gold standards created based on DBpedia, we obtain encouraging results from a thorough evaluation involving four baseline similarity measures and over 15 comparative models based on variants of the proposed method. The proposed method makes significant improvement over baseline models in terms of F1 measure (mostly between 7% and 40%), and it always scores the highest precision and is also among the top performers in terms of recall. We also make public the datasets used in this work, which we believe make the largest collection of gold standards for evaluating relation alignment in the LOD context.

  • 5014.
    Zhang, Zuotai
    et al.
    KTH, School of Industrial Engineering and Management (ITM), Materials Science and Engineering.
    Seetharaman, Seshadri
    KTH, School of Industrial Engineering and Management (ITM), Materials Science and Engineering.
    Li, Wenchao
    Wang, Xidong
    Synthesis and characterization of MgAlON-BN composites2007In: INTERNATIONAL JOURNAL OF MATERIALS RESEARCH, ISSN 1862-5282, Vol. 98, no 1, 64-71 p.Article in journal (Refereed)
    Abstract [en]

    In the present paper the Gibbs energy of formation of MgAlON was evaluated, and on this basis the phase stability diagram of Mg-Al-O-N-B was established. Dense magnesium aluminium oxynitride-boron nitride (MgAlON-BN) composites with 0-30 vol.% BN were synthesized by hot pressing in the temperature range 1750-1850 degrees C in the phase stability region. The phase compositions of the composites analyzed by X-ray diffraction indicated that the main phases were MgAlON and h-BN and no impurities were found. The microstructures of the composites analyzed by transmission electron microscopy, scanning electron microscopy, and high-resolution electron microscopy showed that the MgAlON did not react with BN, and the latter was distributed on the grain boundary of MgAlON homogeneously. Excessive BN (> 20 vol.%) resulted in a discontinuous microstructure for MgAlON. A matrix-flushing method was employed in the quantitative X-ray diffraction analysis for the multi-component MgAlON-BN composites. The results showed that the relative MgAlON content decreased in the process of sintering MgAlON-BN composites. Thermodynamic analysis showed that some Al2O3, AlN, and MgO escaped under the reaction conditions. The lattice parameters of MgAlON in the composites have also been evaluated.

  • 5015.
    Zhang, Zuotai
    et al.
    KTH, School of Industrial Engineering and Management (ITM), Materials Science and Engineering.
    Teng, Li-Dong
    KTH, School of Industrial Engineering and Management (ITM), Materials Science and Engineering.
    Li, Wenchao
    Mechanical Properties and Microstructures of hot-pressed MgAlON-BN Composites2007In: Journal of the European Ceramic Society, ISSN 0955-2219, E-ISSN 1873-619X, Vol. 27, no 1, 319-326 p.Article in journal (Refereed)
    Abstract [en]

    The relationship between the mechanical properties and microstructure of hot-pressed MgAlON-BN composite materials was investigated by scanning electron microscopy (SEM), transmission electron microscopy (TEM), high resolution electron microscopy (HREM) and X-ray diffraction (XRD) techniques. The phase compositions of hot-pressed samples prepared from starting mixtures of Al2O3, AlN, MgO and h-BN consisted of MgAlON phases as a matrix and BN phases as the second phase. The density, bending strength at room temperature, fracture toughness and Vickers hardness were measured. The results indicated that the density, strength and Vickers hardness decrease with increasing h-BN content due to the non-reactive nature and layered structure of h-BN. The fracture toughness, however increased with increasing h-BN addition, reaching a maximum of 3.64 MPa m(0.5); it decreased with further increase of BN content. The increase of fracture toughness was attributed to the presence of microcracks and the decrease was considered to be the discontinuous microstructure of the MgAlON phases. Temperature dependence of bending strength remained constant at low temperature, followed by an increase at 800 degrees C and then, dropped quickly. The increase in the bending strength of the composite was attributed to the decrease of residual stress and to the interwoven microstructure of the composites which prevented grain boundary slip and reduced the attenuation rate of high temperature strength. The machinability of the composites was examined. The results indicate that the composite materials with BN content more than 15 vol.% exhibit excellent machinability and could be drilled using conventional hard metal alloy drills

  • 5016.
    Zhao, Jing
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Temporal weighting of clinical events in electronic health records for pharmacovigilance2015In: 2015 IEEE International Conference on Bioinformatics and Biomedicine: Proceedings / [ed] Jun (Luke) Huan et al., IEEE Computer Society, 2015, 375-381 p.Conference paper (Refereed)
    Abstract [en]

    Electronic health records (EHRs) have recently been identified as a potentially valuable source for monitoring adverse drug events (ADEs). However, ADEs are heavily under- reported in EHRs. Using machine learning algorithms to automatically detect patients that should have had ADEs reported in their health records is an efficient and effective solution. One of the challenges to that end is how to take into account temporality when using clinical events, which are time stamped in EHRs, as features for machine learning algorithms to exploit. Previous research on this topic suggests that representing EHR data as a bag of temporally weighted clinical events is promising; however, how to assign weights in an optimal manner remains unexplored. In this study, nine different temporal weighting strategies are proposed and evaluated using data extracted from a Swedish EHR database, where the predictive performance of models constructed with the random forest learning algorithm is compared. Moreover, variable importance is analyzed to obtain a deeper understanding as to why a certain weighting strategy is favored over another, as well as which clinical events undergo the biggest changes in importance with the various weighting strategies. The results show that the choice of weighting strategy has a significant impact on the predictive performance for ADE detection, and that the best choice of weighting strategy depends on the target ADE and, specifically, on its dose-dependency.

  • 5017.
    Zhao, Jing
    et al.
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Henriksson, Aron
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Learning temporal weights of clinical events using variable importance2016In: BMC Medical Informatics and Decision Making, ISSN 1472-6947, E-ISSN 1472-6947, Vol. 16, no Suppl. 2, 71Article in journal (Refereed)
    Abstract [en]

    Background: Longitudinal data sources, such as electronic health records (EHRs), are very valuable for monitoring adverse drug events (ADEs). However, ADEs are heavily under-reported in EHRs. Using machine learning algorithms to automatically detect patients that should have had ADEs reported in their health records is an efficient and effective solution. One of the challenges to that end is how to take into account the temporality of clinical events, which are time stamped in EHRs, and providing these as features for machine learning algorithms to exploit. Previous research on this topic suggests that representing EHR data as a bag of temporally weighted clinical events is promising; however, the weights were in that case pre-assigned according to their time stamps, which is limited and potentially less accurate. This study therefore focuses on how to learn weights that effectively take into account the temporality and importance of clinical events for ADE detection. Methods: Variable importance obtained from the random forest learning algorithm is used for extracting temporal weights. Two strategies are proposed for applying the learned weights: weighted aggregation and weighted sampling. The first strategy aggregates the weighted clinical events from different time windows to form new features; the second strategy retains the original features but samples them by using their weights as probabilities when building each tree in the forest. The predictive performance of random forest models using the learned weights with the two strategies is compared to using pre-assigned weights. In addition, to assess the sensitivity of the weight-learning procedure, weights from different granularity levels are evaluated and compared. Results: In the weighted sampling strategy, using learned weights significantly improves the predictive performance, in comparison to using pre-assigned weights; however, there is no significant difference between them in the weighted aggregation strategy. Moreover, the granularity of the weight learning procedure has a significant impact on the former, but not on the latter. Conclusions: Learning temporal weights is significantly beneficial in terms of predictive performance with the weighted sampling strategy. Moreover, weighted aggregation generally diminishes the impact of temporal weighting of the clinical events, irrespective of whether the weights are pre-assigned or learned.

  • 5018.
    Zhao, Jing
    et al.
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Henriksson, Aron
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Asker, Lars
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Boström, Henrik
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Detecting Adverse Drug Events with Multiple Representations of Clinical Measurements2014In: 2014 IEEE International Conference on Bioinformatics and Biomedicine (BIBM): Proceedings, IEEE Computer Society, 2014, 536-543 p.Conference paper (Refereed)
    Abstract [en]

    Adverse drug events (ADEs) are grossly under-reported in electronic health records (EHRs). This could be mitigated by methods that are able to detect ADEs in EHRs, thereby allowing for missing ADE-specific diagnosis codes to be identified and added. A crucial aspect of constructing such systems is to find proper representations of the data in order to allow the predictive modeling to be as accurate as possible. One category of EHR data that can be used as indicators of ADEs are clinical measurements. However, using clinical measurements as features is not unproblematic due to the high rate of missing values and they can be repeated a variable number of times in each patient health record. In this study, five basic representations of clinical measurements are proposed and evaluated to handle these two problems. An empirical investigation using random forest on 27 datasets from a real EHR database with different ADE targets is presented, demonstrating that the predictive performance, in terms of accuracy and area under ROC curve, is higher when representing clinical measurements crudely as whether they were taken or how many times they were taken by a patient. Furthermore, a sixth alternative, combining all five basic representations, significantly outperforms using any of the basic representation except for one. A subsequent analysis of variable importance is also conducted with this fused feature set, showing that when clinical measurements have a high missing rate, the number of times they were taken by one patient is ranked as more informative than looking at their actual values. The observation from random forest is also confirmed empirically using other commonly employed classifiers. This study demonstrates that the way in which clinical measurements from EHRs are presented has a high impact for ADE detection, and that using multiple representations outperforms using a basic representation.

  • 5019.
    Zhao, Jing
    et al.
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Henriksson, Aron
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Asker, Lars
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Boström, Henrik
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Detecting Adverse Drug Events with Multiple Representations of Clinical Measurements2014In: Proceedings - 2014 IEEE International Conference on Bioinformatics and Biomedicine, IEEE BIBM 2014, IEEE Computer Society, 2014, 536-543 p., 6999216Conference paper (Refereed)
    Abstract [en]

    Adverse drug events (ADEs) are grossly under-reported in electronic health records (EHRs). This could be mitigated by methods that are able to detect ADEs in EHRs, thereby allowing for missing ADE-specific diagnosis codes to be identified and added. A crucial aspect of constructing such systems is to find proper representations of the data in order to allow the predictive modeling to be as accurate as possible. One category of EHR data that can be used as indicators of ADEs are clinical measurements. However, using clinical measurements as features is not unproblematic due to the high rate of missing values and they can be repeated a variable number of times in each patient health record. In this study, five basic representations of clinical measurements are proposed and evaluated to handle these two problems. An empirical investigation using random forest on 27 datasets from a real EHR database with different ADE targets is presented, demonstrating that the predictive performance, in terms of accuracy and area under ROC curve, is higher when representing clinical measurements crudely as whether they were taken or how many times they were taken by a patient. Furthermore, a sixth alternative, combining all five basic representations, significantly outperforms using any of the basic representation except for one. A subsequent analysis of variable importance is also conducted with this fused feature set, showing that when clinical measurements have a high missing rate, the number of times they were taken by one patient is ranked as more informative than looking at their actual values. The observation from random forest is also confirmed empirically using other commonly employed classifiers. This study demonstrates that the way in which clinical measurements from EHRs are presented has a high impact for ADE detection, and that using multiple representations outperforms using a basic representation.

  • 5020.
    Zhao, Jing
    et al.
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Henriksson, Aron
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Asker, Lars
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Boström, Henrik
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Predictive modeling of structured electronic health records for adverse drug event detection2015In: BMC Medical Informatics and Decision Making, ISSN 1472-6947, E-ISSN 1472-6947, Vol. 15, no SIArticle in journal (Refereed)
    Abstract [en]

    Background: The digitization of healthcare data, resulting from the increasingly widespread adoption of electronic health records, has greatly facilitated its analysis by computational methods and thereby enabled large-scale secondary use thereof. This can be exploited to support public health activities such as pharmacovigilance, wherein the safety of drugs is monitored to inform regulatory decisions about sustained use. To that end, electronic health records have emerged as a potentially valuable data source, providing access to longitudinal observations of patient treatment and drug use. A nascent line of research concerns predictive modeling of healthcare data for the automatic detection of adverse drug events, which presents its own set of challenges: it is not yet clear how to represent the heterogeneous data types in a manner conducive to learning high-performing machine learning models. Methods: Datasets from an electronic health record database are used for learning predictive models with the purpose of detecting adverse drug events. The use and representation of two data types, as well as their combination, are studied: clinical codes, describing prescribed drugs and assigned diagnoses, and measurements. Feature selection is conducted on the various types of data to reduce dimensionality and sparsity, while allowing for an in-depth feature analysis of the usefulness of each data type and representation. Results: Within each data type, combining multiple representations yields better predictive performance compared to using any single representation. The use of clinical codes for adverse drug event detection significantly outperforms the use of measurements; however, there is no significant difference over datasets between using only clinical codes and their combination with measurements. For certain adverse drug events, the combination does, however, outperform using only clinical codes. Feature selection leads to increased predictive performance for both data types, in isolation and combined. Conclusions: We have demonstrated how machine learning can be applied to electronic health records for the purpose of detecting adverse drug events and proposed solutions to some of the challenges this presents, including how to represent the various data types. Overall, clinical codes are more useful than measurements and, in specific cases, it is beneficial to combine the two.

  • 5021.
    Zhao, Jing
    et al.
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Henriksson, Aron
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Asker, Lars
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Boström, Henrik
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Predictive modeling of structured electronic health records for adverse drug event detection2015In: BMC Medical Informatics and Decision Making, ISSN 1472-6947, E-ISSN 1472-6947, Vol. 15, no 4, S1Article in journal (Refereed)
    Abstract [en]

    Background: The digitization of healthcare data, resulting from the increasingly widespread adoption of electronic health records, has greatly facilitated its analysis by computational methods and thereby enabled large-scale secondary use thereof. This can be exploited to support public health activities such as pharmacovigilance, wherein the safety of drugs is monitored to inform regulatory decisions about sustained use. To that end, electronic health records have emerged as a potentially valuable data source, providing access to longitudinal observations of patient treatment and drug use. A nascent line of research concerns predictive modeling of healthcare data for the automatic detection of adverse drug events, which presents its own set of challenges: it is not yet clear how to represent the heterogeneous data types in a manner conducive to learning high-performing machine learning models. Methods: Datasets from an electronic health record database are used for learning predictive models with the purpose of detecting adverse drug events. The use and representation of two data types, as well as their combination, are studied: clinical codes, describing prescribed drugs and assigned diagnoses, and measurements. Feature selection is conducted on the various types of data to reduce dimensionality and sparsity, while allowing for an in-depth feature analysis of the usefulness of each data type and representation. Results: Within each data type, combining multiple representations yields better predictive performance compared to using any single representation. The use of clinical codes for adverse drug event detection significantly outperforms the use of measurements; however, there is no significant difference over datasets between using only clinical codes and their combination with measurements. For certain adverse drug events, the combination does, however, outperform using only clinical codes. Feature selection leads to increased predictive performance for both data types, in isolation and combined. Conclusions: We have demonstrated how machine learning can be applied to electronic health records for the purpose of detecting adverse drug events and proposed solutions to some of the challenges this presents, including how to represent the various data types. Overall, clinical codes are more useful than measurements and, in specific cases, it is beneficial to combine the two.

  • 5022.
    Zhao, Jing
    et al.
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Henriksson, Aron
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Boström, Henrik
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Cascading Adverse Drug Event Detection in Electronic Health Records2015In: 2015 IEEE International Conference on Data Science and Advanced Analytics (DSAA): Proceedings, IEEE Computer Society, 2015Conference paper (Refereed)
    Abstract [en]

    The ability to detect adverse drug events (ADEs) in electronic health records (EHRs) is useful in many medical applications, such as alerting systems that indicate when an ADE-specific diagnosis code should be assigned. Automating the detection of ADEs can be attempted by applying machine learning to existing, labeled EHR data. How to do this in an effective manner is, however, an open question. The issues addressed in this study concern the granularity of the classification task: (1) If we wish to predict the occurrence of ADE, is it advantageous to conflate the various ADE class labels prior to learning, or should they be merged post prediction? (2) If we wish to predict a family of ADEs or even a specific ADE, can the predictive performance be enhanced by dividing the classification task into a cascading scheme: predicting first, on a coarse level, whether there is an ADE or not, and, in the former case, followed by a more specific prediction on which family the ADE belongs to, and then finally a prediction on the specific ADE within that particular family? In this study, we conduct a series of experiments using a real, clinical dataset comprising healthcare episodes that have been assigned one of eight ADE-related diagnosis codes and a set of randomly extracted episodes that have not been assigned any ADE code. It is shown that, when distinguishing between ADEs and non-ADEs, merging the various ADE labels prior to learning leads to significantly higher predictive performance in terms of accuracy and area under ROC curve. A cascade of random forests is moreover constructed to determine either the family of ADEs or the specific class label; here, the performance is indeed enhanced compared to directly employing a one-step prediction. This study concludes that, if predictive performance is of primary importance, the cascading scheme should be the recommended approach over employing a one-step prediction for detecting ADEs in EHRs.

  • 5023.
    Zhao, Jing
    et al.
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Henriksson, Aron
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Boström, Henrik
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Cascading Adverse Drug Event Detection in Electronic Health Records2015In: 2015 IEEE International Conference on Data Science and Advanced Analytics (DSAA): Proceedings, IEEE Computer Society, 2015, 810-817 p., 7344869Conference paper (Refereed)
    Abstract [en]

    The ability to detect adverse drug events (ADEs) in electronic health records (EHRs) is useful in many medical applications, such as alerting systems that indicate when an ADE-specific diagnosis code should be assigned. Automating the detection of ADEs can be attempted by applying machine learning to existing, labeled EHR data. How to do this in an effective manner is, however, an open question. The issues addressed in this study concern the granularity of the classification task: (1) If we wish to predict the occurrence of ADE, is it advantageous to conflate the various ADE class labels prior to learning, or should they be merged post prediction? (2) If we wish to predict a family of ADEs or even a specific ADE, can the predictive performance be enhanced by dividing the classification task into a cascading scheme: predicting first, on a coarse level, whether there is an ADE or not, and, in the former case, followed by a more specific prediction on which family the ADE belongs to, and then finally a prediction on the specific ADE within that particular family? In this study, we conduct a series of experiments using a real, clinical dataset comprising healthcare episodes that have been assigned one of eight ADE-related diagnosis codes and a set of randomly extracted episodes that have not been assigned any ADE code. It is shown that, when distinguishing between ADEs and non-ADEs, merging the various ADE labels prior to learning leads to significantly higher predictive performance in terms of accuracy and area under ROC curve. A cascade of random forests is moreover constructed to determine either the family of ADEs or the specific class label; here, the performance is indeed enhanced compared to directly employing a one-step prediction. This study concludes that, if predictive performance is of primary importance, the cascading scheme should be the recommended approach over employing a one-step prediction for detecting ADEs in EHRs.

  • 5024.
    Zhao, Jing
    et al.
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Henriksson, Aron
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Boström, Henrik
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Detecting Adverse Drug Events Using Concept Hierarchies of Clinical Codes2014In: 2014 IEEE International Conference on Healthcare Informatics: Proceedings, IEEE Computer Society, 2014, 285-293 p.Conference paper (Refereed)
    Abstract [en]

    Electronic health records (EHRs) provide a potentially valuable source of information for pharmacovigilance. However, adverse drug events (ADEs), which can be encoded in EHRs with specific diagnosis codes, are heavily under-reported. To provide more accurate estimates for drug safety surveillance, machine learning systems that are able to detect ADEs could be used to identify and suggest missing ADE-specific diagnosis codes. A fundamental consideration when building such systems is how to represent the EHR data to allow for accurate predictive modeling. In this study, two types of clinical code are used to represent drugs and diagnoses: the Anatomical Therapeutic Chemical Classification System (ATC) and the International Statistical Classification of Diseases and Health Problems (ICD). More specifically, it is investigated whether their hierarchical structure can be exploited to improve predictive performance. The use of random forests with feature sets that include only the original, low-level, codes is compared to using random forests with feature sets that contain all levels in the hierarchies. An empirical investigation using thirty datasets with different ADE targets is presented, demonstrating that the predictive performance, in terms of accuracy and area under ROC curve, can be significantly improved by exploiting codes on all levels in the hierarchies, compared to using only the low-level encoding. A further analysis is presented in which two strategies are employed for adding features level-wise according to the concept hierarchies: top-down, starting with the highest abstraction levels, and bottom-up, starting with the most specific encoding. The main finding from this subsequent analysis is that predictive performance can be kept at a high level even without employing the more specific levels in the concept hierarchies.

  • 5025.
    Zhao, Jing
    et al.
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Henriksson, Aron
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Boström, Henrik
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Detecting Adverse Drug Events Using Concept Hierarchies of Clinical Codes2014In: 2014 IEEE International Conference on Healthcare Informatics: Proceedings, IEEE Computer Society, 2014, 285-293 p., 7052501Conference paper (Refereed)
    Abstract [en]

    Electronic health records (EHRs) provide a potentially valuable source of information for pharmacovigilance. However, adverse drug events (ADEs), which can be encoded in EHRs with specific diagnosis codes, are heavily under-reported. To provide more accurate estimates for drug safety surveillance, machine learning systems that are able to detect ADEs could be used to identify and suggest missing ADE-specific diagnosis codes. A fundamental consideration when building such systems is how to represent the EHR data to allow for accurate predictive modeling. In this study, two types of clinical code are used to represent drugs and diagnoses: the Anatomical Therapeutic Chemical Classification System (ATC) and the International Statistical Classification of Diseases and Health Problems (ICD). More specifically, it is investigated whether their hierarchical structure can be exploited to improve predictive performance. The use of random forests with feature sets that include only the original, low-level, codes is compared to using random forests with feature sets that contain all levels in the hierarchies. An empirical investigation using thirty datasets with different ADE targets is presented, demonstrating that the predictive performance, in terms of accuracy and area under ROC curve, can be significantly improved by exploiting codes on all levels in the hierarchies, compared to using only the low-level encoding. A further analysis is presented in which two strategies are employed for adding features level-wise according to the concept hierarchies: top-down, starting with the highest abstraction levels, and bottom-up, starting with the most specific encoding. The main finding from this subsequent analysis is that predictive performance can be kept at a high level even without employing the more specific levels in the concept hierarchies.

  • 5026.
    Zhao, Jing
    et al.
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Henriksson, Aron
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Kvist, Maria
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences. Karolinska Institute, Sweden.
    Asker, Lars
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Boström, Henrik
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Handling Temporality of Clinical Events for Drug Safety Surveillance2015In: AMIA Annual Symposium Proceedings, ISSN 1559-4076, Vol. 2015, 1371-1380 p.Article in journal (Refereed)
    Abstract [en]

    Using longitudinal data in electronic health records (EHRs) for post-marketing adverse drug event (ADE) detection allows for monitoring patients throughout their medical history. Machine learning methods have been shown to be efficient and effective in screening health records and detecting ADEs. How best to exploit historical data, as encoded by clinical events in EHRs is, however, not very well understood. In this study, three strategies for handling temporality of clinical events are proposed and evaluated using an EHR database from Stockholm, Sweden. The random forest learning algorithm is applied to predict fourteen ADEs using clinical events collected from different lengths of patient history. The results show that, in general, including longer patient history leads to improved predictive performance, and that assigning weights to events according to time distance from the ADE yields the biggest improvement.

  • 5027.
    Zhao, Jing
    et al.
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Henriksson, Aron
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Kvist, Maria
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Asker, Lars
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Boström, Henrik
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Handling Temporality of Clinical Events for Drug Safety Surveillance2015In: AMIA Annual Symposium Proceedings, ISSN 1559-4076, Vol. 2015, 1371-1380 p.Article in journal (Refereed)
    Abstract [en]

    Using longitudinal data in electronic health records (EHRs) for post-marketing adverse drug event (ADE) detection allows for monitoring patients throughout their medical history. Machine learning methods have been shown to be efficient and effective in screening health records and detecting ADEs. How best to exploit historical data, as encoded by clinical events in EHRs is, however, not very well understood. In this study, three strategies for handling temporality of clinical events are proposed and evaluated using an EHR database from Stockholm, Sweden. The random forest learning algorithm is applied to predict fourteen ADEs using clinical events collected from different lengths of patient history. The results show that, in general, including longer patient history leads to improved predictive performance, and that assigning weights to events according to time distance from the ADE yields the biggest improvement.

  • 5028.
    Zhao, Jing
    et al.
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Karlsson, Isak
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Asker, Lars
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Boström, Henrik
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Applying Methods for Signal Detection in Spontaneous Reports to Electronic Patient Records2013In: Proceedings of the  19th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Association for Computing Machinery (ACM), 2013Conference paper (Refereed)
    Abstract [en]

    Currently, pharmacovigilance relies mainly on disproportionality analysis of spontaneous reports. However, the analysis of spontaneous reports is concerned with several problems, such as reliability, under-reporting and insucient patient information. Longitudinal healthcare data, such as Electronic Patient Records (EPRs) in which comprehensive information of each patient is covered, is a complementary source of information to detect Adverse Drug Events (ADEs). A wide set of disproportionality methods has been developed for analyzing spontaneous reports to assess the risk of reported events being ADEs. This study aims to investigate the use of such methods for detecting ADEs when analyzing EPRs. The data used in this study was extracted from Stockholm EPR Corpus. Four disproportionality methods (proportional reporting rate, reporting odds ratio, Bayesian condence propagation neural network, and Gamma-Poisson shrinker) were applied in two dierent ways to analyze EPRs: creating pseudo spontaneous reports based on all observed drug-event pairs (event-level analysis) or analyzing distinct patients who experienced a drug-event pair (patient-level analysis). The methods were evaluated in a case study on safety surveillance of Celecoxib. The results showed that, among the top 200 signals, more ADEs were detected by the event-level analysis than by the patient-level analysis. Moreover, the event-level analysis also resulted in a higher mean average precision. The main conclusion of this study is that the way in which the disproportionality analysis is applied, the event-level or patient-level analysis, can have a much higher impact on the performance than which disproportionality method is employed.

  • 5029.
    Zhao, Jing
    et al.
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Karlsson, Isak
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Asker, Lars
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Boström, Henrik
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Applying Methods for Signal Detection in Spontaneous Reports to Electronic Patient Records2013In: Proceedings of the  19th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Association for Computing Machinery (ACM), 2013Conference paper (Refereed)
    Abstract [en]

    Currently, pharmacovigilance relies mainly on disproportionality analysis of spontaneous reports. However, the analysis of spontaneous reports is concerned with several problems, such as reliability, under-reporting and insucient patient information. Longitudinal healthcare data, such as Electronic Patient Records (EPRs) in which comprehensive information of each patient is covered, is a complementary source of information to detect Adverse Drug Events (ADEs). A wide set of disproportionality methods has been developed for analyzing spontaneous reports to assess the risk of reported events being ADEs. This study aims to investigate the use of such methods for detecting ADEs when analyzing EPRs. The data used in this study was extracted from Stockholm EPR Corpus. Four disproportionality methods (proportional reporting rate, reporting odds ratio, Bayesian condence propagation neural network, and Gamma-Poisson shrinker) were applied in two dierent ways to analyze EPRs: creating pseudo spontaneous reports based on all observed drug-event pairs (event-level analysis) or analyzing distinct patients who experienced a drug-event pair (patient-level analysis). The methods were evaluated in a case study on safety surveillance of Celecoxib. The results showed that, among the top 200 signals, more ADEs were detected by the event-level analysis than by the patient-level analysis. Moreover, the event-level analysis also resulted in a higher mean average precision. The main conclusion of this study is that the way in which the disproportionality analysis is applied, the event-level or patient-level analysis, can have a much higher impact on the performance than which disproportionality method is employed.

  • 5030.
    Zhao, Jing
    et al.
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Papapetrou, Panagiotis
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Asker, Lars
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Boström, Henrik
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Learning from heterogeneous temporal data from electronic health records2017In: Journal of Biomedical Informatics, ISSN 1532-0464, E-ISSN 1532-0480, Vol. 65, 105-119 p.Article in journal (Refereed)
    Abstract [en]

    Electronic health records contain large amounts of longitudinal data that are valuable for biomedical informatics research. The application of machine learning is a promising alternative to manual analysis of such data. However, the complex structure of the data, which includes clinical events that are unevenly distributed over time, poses a challenge for standard learning algorithms. Some approaches to modeling temporal data rely on extracting single values from time series; however, this leads to the loss of potentially valuable sequential information. How to better account for the temporality of clinical data, hence, remains an important research question. In this study, novel representations of temporal data in electronic health records are explored. These representations retain the sequential information, and are directly compatible with standard machine learning algorithms. The explored methods are based on symbolic sequence representations of time series data, which are utilized in a number of different ways. An empirical investigation, using 19 datasets comprising clinical measurements observed over time from a real database of electronic health records, shows that using a distance measure to random subsequences leads to substantial improvements in predictive performance compared to using the original sequences or clustering the sequences. Evidence is moreover provided on the quality of the symbolic sequence representation by comparing it to sequences that are generated using domain knowledge by clinical experts. The proposed method creates representations that better account for the temporality of clinical events, which is often key to prediction tasks in the biomedical domain.

  • 5031.
    Zhao, Jing
    et al.
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Papapetrou, Panagiotis
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Asker, Lars
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Boström, Henrik
    Stockholms universitet, Institutionen för data- och systemvetenskap.
    Learning from heterogeneous temporal data from electronic health records2017In: Journal of Biomedical Informatics, ISSN 1532-0464, E-ISSN 1532-0480, Vol. 65, 105-119 p.Article in journal (Refereed)
    Abstract [en]

    Electronic health records contain large amounts of longitudinal data that are valuable for biomedical informatics research. The application of machine learning is a promising alternative to manual analysis of such data. However, the complex structure of the data, which includes clinical events that are unevenly distributed over time, poses a challenge for standard learning algorithms. Some approaches to modeling temporal data rely on extracting single values from time series; however, this leads to the loss of potentially valuable sequential information. How to better account for the temporality of clinical data, hence, remains an important research question. In this study, novel representations of temporal data in electronic health records are explored. These representations retain the sequential information, and are directly compatible with standard machine learning algorithms. The explored methods are based on symbolic sequence representations of time series data, which are utilized in a number of different ways. An empirical investigation, using 19 datasets comprising clinical measurements observed over time from a real database of electronic health records, shows that using a distance measure to random subsequences leads to substantial improvements in predictive performance compared to using the original sequences or clustering the sequences. Evidence is moreover provided on the quality of the symbolic sequence representation by comparing it to sequences that are generated using domain knowledge by clinical experts. The proposed method creates representations that better account for the temporality of clinical events, which is often key to prediction tasks in the biomedical domain.

  • 5032. Zhong, Jianghua
    et al.
    Cheng, Daizhan
    Hu, Xiaoming
    KTH, School of Engineering Sciences (SCI), Mathematics (Dept.), Optimization and Systems Theory. KTH, School of Computer Science and Communication (CSC), Centres, Centre for Autonomous Systems, CAS. KTH, School of Electrical Engineering (EES), Centres, ACCESS Linnaeus Centre.
    Constructive stabilization for quadratic input nonlinear systems2008In: Automatica, ISSN 0005-1098, E-ISSN 1873-2836, Vol. 44, no 8, 1996-2005 p.Article in journal (Refereed)
    Abstract [en]

    In this paper stabilization of nonlinear systems with quadratic multi-input is considered. With the help of control Lyapunov function (CLF), a constructive parameterization of controls that globally asymptotically stabilize the system is proposed. Two different cases are considered. Firstly, under certain regularity assumptions. the feasible control set is parameterized, and Continuous feedback stabilizing controls are designed. Then for the general case. piecewise Continuous stabilizing controls are proposed. The design procedure can also be used to verify whether a candidate CLF is indeed a CLF. Several illustrative examples are presented as well.

  • 5033.
    Zhu, Rui
    et al.
    KTH, School of Architecture and the Built Environment (ABE), Urban Planning and Environment, Geodesy and Geoinformatics.
    Gidofalvi, Gyözö
    KTH, School of Architecture and the Built Environment (ABE), Urban Planning and Environment, Geodesy and Geoinformatics.
    GPS-based Crowd Sourced Intelligent Traffic Information Hub2013In: Proceedings of the 26th International Cartographic Conference / [ed] Manfred F. Buchroithner, ICC International Cartographic Association , 2013, 669-670 p.Conference paper (Refereed)
    Abstract [en]

    Congestion is a major problem in most metropolitan areas and given the increasing rate of urbanization it is likely to be an even more serious problem in the rapidly expanding mega cities. Some well know negative effects of congestion include: 1) the economic losses and quality of life degradation that result from the increased and unpredictable travel times, 2) the increased level of carbon footprint that vehicles idling in congestions leave behind, and 3) the increased number of traffic accidents that are direct results of the stress and fatigue of drivers that are stuck in congestion.

    One possible method to combat congestion is provide intelligent traffic management systems that can in a timely manner inform drivers about current or predicted traffic congestion that is relevant to them on their journeys. This this extent, the present paper proposes a scalable, grid-based intelligent traffic information hub that facilitates the manual definition and/or automatic detection of abnormal traffic condition events, e.g., accidents or congestion, and in advance informs drivers about events that will likely be relevant to them on their journey, thereby allowing the divers or their onboard navigation units to alter their paths as needed.

    The proposed system achieves the above described functionality through the following methodology. The system, without loss of generality, adopts a grid-based discretization of space, which by changing the resolution of the grid allows the system to scale in terms of it computation cost and the geographical level of detail of traffic information that it manages. The system derives traffic information from the continuous stream of grid-based position and speed reports that it receives from the vehicles. In particular, the system in an online fashion 1) summarizes Current (grid-based) Traffic Flow Statistics (CTFS), i.e., it records for each grid cell g from each neighboring grid cell n, the mean and standard deviation of the speeds of the vehicles that are currently located in g and have entered g from n; and 2) efficiently incorporates the CTFS into compressed Historical (grid-based) Traffic Flow Statistics (HTFS) using incremental statistics. Simultaneously, using a sliding window model, the system also 1) maintains the Recent (grid-based) Trajectories (RT) of the vehicles; 2) extracts Recent (gridbased) Mobility Statistics (RMS), i.e., it records for each destination grid cell d, for each neighboring grid cell n of g, and for each possible source grid cell s, the number of vehicles that (i) are currently in d, (ii) have entered d from n, and (iii) have a RT that has passed through s; and 3) efficiently incorporates the RMS into compressed Historical (grid-based) Mobility Statistics (HMS) using incremental statistics. To capture the temporal variability in traffic flow and mobility patterns at different scales, the system through temporal domain projections maintains day-of-week and hour-ofday based aggregations of HTFS and RMS. Then, the system classifies a grid cell g to be congested from the direction of a neighboring grid cell n if the current mean speed of vehicles that entered the grid cell g from the direction of n is below the normal according to the temporally relevant HFS. Finally, based on the temporally relevant HMS, the system sends out congestion notifications to vehicles that are likely to be effected in the future part of their journey by these congestions, i.e., the system sends out a congestion notification (g,n) to a vehicle v that is currently located in some grid cell s from which the likelihood of v moving to g through n within the prediction horizon is above a user-defined threshold.

    Extensive empirical evaluations on large sets of realistically simulated trajectories of vehicles illustrate that the above described methodology and its simple SQL-based implementation in a relational database system is scalable and effective. In particular, the execution time of- and the space used by the system scales linearly with the input size (number of concurrently moving vehicles) and the method’s mutually dependent parameters (grid resolution r and RT length l) that jointly define a spatio-temporal resolution. Within the area of a large city (40km by 40km), assuming a 60km/h average vehicle speed, the system, running on a single personal computer, can manage the described congestion detection and one-minute-ahead notification tasks within real-time requirements for 15 thousand and 2.5 million concurrently moving vehicles for spatio-temporal resolutions (r=62.5m, l=17) and (r=4km,l=2), respectively. Finally, the proposed method, for all spatio-temporal resolutions and prediction horizons, significantly outperforms in terms of notification accuracy the grid-based baseline method, which sends non-directional congestion notifications based on the recent linear movement tendencies of vehicles. 

  • 5034.
    Ziemke, Tom
    University of Skövde, School of Humanities and Informatics.
    Adaptive Behavior in Autonomous Agents1998Report (Other academic)
    Abstract [en]

    This paper gives an overview of the bottom-up approach to artificial intelligence (AI), commonly referred to as behavior-oriented AI. The behavior-oriented approach, with its focus on the interaction between autonomous agents and their environments, is introduced by contrasting it with the traditional approach of knowledge-based AI. Different notions of autonomy are discussed, and key problems of generating adaptive and complex behavior are identified. A number of techniques for the generation of behavior are introduced and evaluated regarding their potential for realizing different aspects of autonomy as well as adaptivity and complexity of behavior. It is concluded that in order to realize truly autonomous and intelligent agents, the behavior-oriented approach will have to focus even more on life-like qualities in both agents and environments.

  • 5035.
    Ziemke, Tom
    University of Skövde, School of Humanities and Informatics.
    Radar Image Segmentation using Recurrent Artificial Neural Networks1996Report (Other academic)
    Abstract [en]

    This paper discusses the application of artificial neural networks to the segmentation of Doppler radar images, in particular the detection of oil spills within sea environments, based on a classification of radar backscatter signals. Best results have been achieved with recurrent backpropagation networks of an architecture similar to that of Elman's Simple Recurrent Network. The recurrent networks are shown to be very robust to variations in both sea state (weather conditions) as well as illumination distance, and their performance is analysed in further detail.

  • 5036.
    Ziemke, Tom
    University of Skövde, School of Humanities and Informatics.
    Radar Image Segmentation using Second-Order Recurrent Networks1996Report (Other academic)
    Abstract [en]

    A second-order recurrent artificial neural network architecture for the segmentation and integration of radar images is introduced in this paper. This architecture consists of two sub-networks: a function network that classifies radar measurements into four different categories of objects in sea environments (water, oil spills, land and boats), and a context network that dynamically computes the function network's input weights. It is shown that this mechanism allows networks to learn to solve the task through a dynamic adaptation of their weighting of different radar measurements.behaviour.

  • 5037.
    Ziemke, Tom
    University of Skövde, School of Humanities and Informatics.
    Radar Image Segmentation using Self-Adapting Recurrent Networks1997Report (Other academic)
    Abstract [en]

    This paper presents a novel approach to the segmentation and integration of (radar) images using a second-order recurrent artificial neural network architecture consisting of two sub- networks: a function network that classifies radar measurements into four different categories of objects in sea environments (water, oil spills, land and boats), and a context network that dynamically computes the function network's input weights. It is shown that in experiments (using simulated radar images) this mechanism outperforms conventional artificial neural networks since it allows the network to learn to solve the task through a dynamic adaptation of its classification function based on its internal state closely reflecting the current context.

  • 5038.
    Ziemke, Tom
    University of Skövde, School of Humanities and Informatics.
    Recurrent Artficial Neural Networks for the Detection of Oil Spills from Doppler Radar Imagery1995Report (Other academic)
    Abstract [en]

    This paper discusses the application of artificial neural networks (ANNs) to the detection of oil spills in sea clutter environments from the classification of radar backscatter signals. A comparison and evaluation of different network architectures regarding reliability of dection and robustness to varying sea states/wind conditions shows that for this problem best results are achieved with a recurrent architecture similar to that of Elman's SRN.

  • 5039.
    Ziemke, Tom
    University of Skövde, School of Humanities and Informatics.
    Remembering how to behave: Recurrent neural networks for adaptive robot behavior.1999Report (Other academic)
    Abstract [en]

    This paper discusses the use of recurrent neural networks for control of and learning in robots and autonomous agents. In particular the use of feedback in both first- and higher-order recurrent network architectures for the realization of adaptive robot behavior is investigated. Two experiments, in which controller network weights are evolved to solve tasks requiring robots to exhibit context- or state-dependent behavior, are used to demonstrate and analyze different recurrent control architectures.

  • 5040.
    Ziemke, Tom
    University of Skövde, School of Humanities and Informatics.
    Rethinking Grounding1999Report (Other academic)
    Abstract [en]

    The grounding problem is, generally speaking, the problem of causally connecting an artificial agent with its environment such that the agent's (internal) mechanisms and behaviour can be intrinsic and meaningful to itself, rather than dependent on an external designer or observer. This paper briefly reviews Searle's and Harnad's analyses of the grounding problem are and evaluates cognitivist and enactivist approaches to solving it. It is argued that, although the two categories of grounding approaches differ in their nature and the problems they have to face, both, so far, fail to provide fully grounded systems. Further it is argued here that the reason the problem is somewhat underestimated lies in the notions of situatedness and embodiment in modern AI, which goes beyond purely computational systems, but fails to acknowledge the historically grounded nature of the relation between living systems and their environments.

  • 5041.
    Ziemke, Tom
    University of Skövde, School of Humanities and Informatics.
    The `Environmental Puppeteer' Revisited: A Connectionist Perspective on Autonomy´.1997Report (Other academic)
    Abstract [en]

    Today's `autonomous´robots only have very limited autonomy and are in fact very much under the control of the `environmental puppeteer', i.e their behaviour is

    determined, via virtual strings, by environmental conditions. Hence, it has been stated as the goal of modern scientific robotics to "cut the strings and give the robot its autonomy''. Different notions of autonomy in artefacts and living systems are examined in this paper, and different aspects/dimensions of autonomy are identified and illustrated with examples from connectionist robot control. A connectionist architecture is introduced that aims to increase robotic autonomy through integration of connectionist self-organisation/learning with the enactive view of structural coupling between environment and agent. In the resulting robot control architecture it is the environment that is pulling the strings, but the agent that develops them and dynamically decides which of them to use in a particular situation. Hence, the notion of autonomy advocated here is not `independence of environment' (a `freedom' most artefacts have), but rather an agent's capacity to actively embed itself in its environment and flexibly utilize it as a resource.

  • 5042.
    Ziemke, Tom
    University of Skövde, School of Humanities and Informatics.
    Towards Adaptive Behaviour System Integration using Connectionist Infinite State Automata1996Report (Other academic)
    Abstract [en]

    A higher order recurrent connectionist architecture for adaptive control of autonomous robots is introduced in this paper. This architecture, inspired by Pollack's Sequential Cascaded Network, consists of two sub-networks: a function network for the coupling between sensory inputs and motor outputs, and a context network, which dynamically adapts the function network in order to allow a flexible mapping from percepts to actions. The approach taken here is compared to dynamics and algorithmic approach to autonomous robot control, and it is argued that the above architecture allows an integration of (a) the complex structure and control typical for the algorithmic approach, (b) the capacity to utilize systematically continuous state spaces, and (c) the self- organizing learning capacity of connectionist systems with a simple, but powerful mechanism for context-dependent adaptation of behaviour.

  • 5043.
    Ziemke, Tom
    University of Skövde, School of Humanities and Informatics.
    Towards Adaptive Perception in Autonomous Robots using Second-Order Recurrent Networks1996Report (Other academic)
    Abstract [en]

    In this paper a higher-order recurrent connectionist architecture is used for learning adaptive behaviour in an autonomous robot. This architecture consists of two sub-networks in a master-slave relationship: a function network for the coupling between sensory inputs and motor outputs, and a context network, which dynamically adapts the sensory input weights in order to allow a flexible, context-dependent mapping from percepts to actions. The capabilities of this architecture are demonstrated in a number of action selection experiments with a simulated Khepera robot, and it is argued that the general approach of generically dividing the overall control task between sequentially cascaded context and function learning offers a powerful mechanism for autonomous long- and short-term adaptation of behaviour.

  • 5044.
    Ziemke, Tom
    University of Skövde, School of Humanities and Informatics.
    Towards Autonomous Robot Control via Self-Adapting Recurrent Networks1996Report (Other academic)
    Abstract [en]

    In this paper a higher-order recurrent connectionist architecture is used for learning adaptive behaviour in an autonomous robot. This architecture consists of two sub-networks in a master-slave relationship: a function network for the coupling between sensory inputs and motor outputs, and a context network, which dynamically adapts the sensory input weights in order to allow a flexible, context-dependent mapping from percepts to actions. The capabilities of this architecture are demonstrated in a number of action selection experiments with a simulated Khepera robot, and it is argued that the general approach of generically dividing the overall control task between sequentially cascaded context and function learning offers a powerful mechanism for autonomous long- and short-term adaptation of behaviour

  • 5045.
    Ziemke, Tom
    et al.
    University of Skövde, School of Humanities and Informatics.
    Athley, Fredrik
    University of Skövde, School of Humanities and Informatics.
    Oil Spill Detection from Doppler Radar Imagery using Artificial Neural Networks1995Report (Other academic)
    Abstract [en]

    This paper reports on results of an ongoing project investigating the application of artificial neural networks (ANNs) to the classification/ cartography of sea clutter environments, and in particular the detection of oil spills, on the basis of their radar backscatter signals.

  • 5046.
    Ziemke, Tom
    et al.
    University of Skövde, School of Humanities and Informatics.
    Atley, Fredrik
    University of Skövde, School of Humanities and Informatics.
    Connectionist Models for the Detection of Oil Spills from Doppler Radar Imagery1995Report (Other academic)
    Abstract [en]

    This paper reports on the results of a project investigating the potential of applying artificial neural networks to the problem of detecting oil spills on basis of the radar backscatter signals from a sea clutter environment illuminated by a Doppler radar. Recurrent backpropagation models which were found to exhibit satisfactory performance, superior to that of feed-forward networks, are discussed and analysed in particular.

  • 5047.
    Ziemke, Tom
    et al.
    University of Skövde, School of Humanities and Informatics.
    Bodén, Mikael
    University of Skövde, School of Humanities and Informatics.
    Niklasson, Lars
    University of Skövde, School of Humanities and Informatics.
    Oil Spill Detection: A Case Study of Recurrent Artificial Neural Networks1997Report (Other academic)
    Abstract [en]

    This paper summarizes and analyzes the results of a case study of artificial neural networks for the detection of oil spills from radar imagery, which has been carried as a joint project between the Connectionist Research Group, University of Skövde, and Ericsson Microwave Systems AB, Mölndal, Sweden.

  • 5048.
    Ziemke, Tom
    et al.
    University of Skövde, School of Humanities and Informatics.
    Jirenhed, Dan-Anders
    University of Skövde, School of Humanities and Informatics.
    Hesslow, Germund
    University of Skövde, School of Humanities and Informatics.
    Blind Adaptive Behavior Based on Internal Simulation of Perception2002Report (Other academic)
    Abstract [en]

    This paper presents experiments, based on a neuroscientific hypothesis, exploring the possibility of an 'inner world' based on internal simulation of perception rather than an explicit representational world model. First a series of initial experiments is discussed, in which recurrent neural networks were evolved to (a) control collision-free corridor following behavior in a simulated Khepera robot, and (b) predict the next time step's sensory input as accurately as possible. Attempts to let the robot act 'blindly', repeatedly using its own prediction instead of the real sensory input, were not particularly successful. This motivated the second series of experiments, on which this paper focuses. A feed-forward network was used which, as above, controlled behavior and predicted sensory input. However, weight evolution was now guided by the sole fitness criterion of successful, 'blind' corridor following behaviour, including timely turns, as above using as input only own predictions rather than real sensory input. The trained robot is in some cases actually able to move 'blindly' in a simple environment for hundreds of time steps, successfully handling several multi-step turns. Somewhat surprisingly, however, it does so based on self-generated input that is very different from the actual sensory values.

  • 5049.
    Zikra, Iyad
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Implementing the Unifying Meta-Model for Enterprise Modeling and Model-Driven Development: An Experience Report2012In: The Practice of Enterprise Modeling: 5th IFIP WG 8.1 Working Conference, PoEM 2012. Proceedings / [ed] Kurt Sandkuhl, Ulf Seigerroth, Janis Stirna, Springer, 2012, 172-187 p.Conference paper (Refereed)
    Abstract [en]

    Model-Driven Development (MDD) is becoming increasingly popular as a choice for developing information systems. Tools that support the principles of MDD are also growing in number and variety of available functionality. MetaEdit+ is a meta-modeling tool used for developing Domain Specific Languages and is identified as an MDD tool. The Eclipse Modeling Framework (EMF) and Graphical Modeling Project (GMP) are two Eclipse projects that provide plug-ins to support the principles of MDD. In this paper, we report on our experience in using MetaEdit+ and the Eclipse plug-ins for developing a graphical editor for the unifying meta-model, which is an MDD approach that extends the traditional view of MDD to cover Enterprise Modeling. The two modeling environments are reviewed using functionality areas that are identified by the research community as necessary in MDD tools. This report will provide useful insights for researchers and practitioners alike concerning the use of MetaEdit+ and the Eclipse plug-ins as MDD tools.

  • 5050.
    Zikra, Iyad
    Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
    Model-based Requirements for Integrating Cloud Services2016In: Proceedings of the CAiSE'16 Forum, at the 28th International Conference on Advanced Information Systems Engineering (CAiSE 2016) / [ed] Sergio España, Mirjana Ivanović, Miloš Savić, 2016, 65-72 p.Conference paper (Refereed)
    Abstract [en]

    Cloud-based services provide an alternative to the in-house implementation of various types of functionality. Organizations rely on such services to minimize the need for long-term commitments and enhance scalability and ubiquitous access to the services. However, achieving complex tasks that require a combination of services is not well studied, despite the potential added value. This paper investigates the requirements encountered when integrating cloud-based services in the modern organization. The paper proposes a model-driven solution for capturing the requirements for integrating cloud-based services. The model is to be used within the larger context of the organizational design; modeling components used to describe requirements are related to other views of the organization. A prototype tool and an example business case are presented to illustrate how the requirements model can be elicited and designed. The models are capable of being transformed into an integration solution.

9899100101102103 5001 - 5050 of 5121
CiteExportLink to result list
Permanent link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf