Multilevel Weighted Support Vector Machine for Classification on Healthcare Data with Missing Values

• Published in: PloS one Vol. 11; no. 5; p. e0155119
• Main Authors: Razzaghi, Talayeh, Roderick, Oleg, Safro, Ilya, Marko, Nicholas
• Format: Journal Article
• Language: English
• Published: United States Public Library of Science 19.05.2016; Public Library of Science (PLoS)
• Subjects: Algorithms; Analysis; Analytics; Artificial intelligence; Benchmarking; Biology and Life Sciences; Classification; Computer and Information Sciences; Computer science; Data mining; Data Mining - methods; Data processing; Databases, Factual; Datasets; Decision making; Electronic Health Records; Electronic medical records; False Positive Reactions; Feasibility studies; Gene expression; Health care; Health care access; Health Services Research; Humans; Informatics; Management; Medical Informatics - instrumentation; Medical Informatics - methods; Medical records; Methods; Models, Theoretical; Multilevel; Patients; Physical Sciences; Preprocessing; Production methods; Regression Analysis; Research and Analysis Methods; Social Sciences; Software; Sparsity; Standard data; Support Vector Machine; Support vector machines; Teaching methods; United States; United States > US
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0155119

This work is motivated by the needs of predictive analytics on healthcare data as represented by Electronic Medical Records. Such data is invariably problematic: noisy, with missing entries, with imbalance in classes of interests, leading to serious bias in predictive modeling. Since standard data mining methods often produce poor performance measures, we argue for development of specialized techniques of data-preprocessing and classification. In this paper, we propose a new method to simultaneously classify large datasets and reduce the effects of missing values. It is based on a multilevel framework of the cost-sensitive SVM and the expected maximization imputation method for missing values, which relies on iterated regression analyses. We compare classification results of multilevel SVM-based algorithms on public benchmark datasets with imbalanced classes and missing values as well as real data in health applications, and show that our multilevel SVM-based method produces fast, and more accurate and robust classification results.