Document Detail

A Pattern Mining Approach for Classifying Multivariate Temporal Data.
MedLine Citation:
PMID:  22267987     Owner:  NLM     Status:  Publisher    
We study the problem of learning classification models from complex multivariate temporal data encountered in electronic health record systems. The challenge is to define a good set of features that are able to represent well the temporal aspect of the data. Our method relies on temporal abstractions and temporal pattern mining to extract the classification features. Temporal pattern mining usually returns a large number of temporal patterns, most of which may be irrelevant to the classification task. To address this problem, we present the minimal predictive temporal patterns framework to generate a small set of predictive and non-spurious patterns. We apply our approach to the real-world clinical task of predicting patients who are at risk of developing heparin induced thrombocytopenia. The results demonstrate the benefit of our approach in learning accurate classifiers, which is a key step for developing intelligent clinical monitoring systems.
Iyad Batal; Hamed Valizadegan; Gregory F Cooper; Milos Hauskrecht
Related Documents :
21784667 - Logistic vs hierarchical modeling: an analysis of a statewide inpatient sample.
22729927 - Principles of multilevel analysis and its relevance to studies of antimicrobial resista...
21931467 - Semiparametric additive risks regression for two-stage design survival studies.
22116777 - Critical aspects of urine profiling for the selection of potential biomarkers using upl...
22935907 - A procedure for assessing intervention fidelity in experiments testing educational and ...
19320707 - The great cannabis classification debacle: what are the likely consequences for policin...
Publication Detail:
Journal Detail:
Title:  Proceedings. IEEE International Conference on Bioinformatics and Biomedicine     Volume:  2011     ISSN:  2156-1125     ISO Abbreviation:  -     Publication Date:  2011 Nov 
Date Detail:
Created Date:  2012-1-23     Completed Date:  -     Revised Date:  -    
Medline Journal Info:
Nlm Unique ID:  101525347     Medline TA:  Proceedings (IEEE Int Conf Bioinformatics Biomed)     Country:  -    
Other Details:
Languages:  ENG     Pagination:  358-365     Citation Subset:  -    
Department of Computer Science University of Pittsburgh.
Export Citation:
APA/MLA Format     Download EndNote     Download BibTex
MeSH Terms
Grant Support
R21 LM009102-02//NLM NIH HHS; R01 GM088224-02//NIGMS NIH HHS; R01 GM088224-03//NIGMS NIH HHS; R01 LM010019-02//NLM NIH HHS; R01 LM010019-03//NLM NIH HHS

From MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine

Previous Document:  Design, Synthesis, and Biological Evaluation of PKD Inhibitors.
Next Document:  Assessing the Diagnostic Imaging needs for Five Selected Hospitals in Uganda.