Stable clinical prediction using graph support vector machines

The stability matters in clinical prediction models because it makes the model to be interpretable and generalizable. It is paramount for high dimensional data, which employ sparse models with feature selection ability. We propose a new method to stabilize sparse support vector machines using intrinsic graph structure of the electronic medical records. The graph structure is exploited using the Jaccard similarity among features. Our method employs a convex function to penalize the pairwise l∞-norm of connected feature coefficients in the graph. We apply the alternating direction method of multipliers to solve the proposed formulation. Our experiments are conducted on a synthetic and three real-world hospital datasets. We show that our proposed method is more stable than the state-of-the-art feature selection and classification techniques in terms of three stability measures namely, Jaccard similarity measure, Spearman's rank correlation coefficient and Kuncheva index. We further show that our method has resulted in better classification performance compared to the baselines.

• Publication year

  2016

• Venue

  International Conference on Pattern Recognition

• Publication date

  Unknown publication date

• Fields of study

  Medicine, Computer Science, Mathematics

• Identifiers
  DOI 10.1109/ICPR.2016.7900148
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Stable feature selection for clinical prediction: Exploiting ICD tree structure using Tree-Lasso

  2015cited by this paper
• Big data: a revolution that will transform how we live, work, and think

  2014cited by this paper
• Improving the electronic health record--are clinicians getting what they wished for?

  2013cited by this paper
• Feature grouping and selection over an undirected graph

  2012cited by this paper
• Mining electronic health records: towards better research applications and clinical care

  2012cited by this paper
• Effect of Size and Heterogeneity of Samples on Biomarker Discovery: Synthetic and Real Data Assessment

  2012influential reference
• Efficient variable selection in support vector machines via the alternating direction method of multipliers

  2011cited by this paper
• Stable feature selection via dense feature groups

  2008cited by this paper
• Simultaneous Regression Shrinkage, Variable Selection, and Supervised Clustering of Predictors with OSCAR

  2008cited by this paper
• An Improved 1-norm SVM for Simultaneous Classification and Variable Selection

  2007cited by this paper
• A stability index for feature selection

  2007cited by this paper
• The doubly regularized support vector machine

  2006influential reference
• Model selection and estimation in regression with grouped variables

  2006influential reference
• On Model Selection Consistency of Lasso

  2006cited by this paper
• Sparsity and smoothness via the fused lasso

  2005influential reference
• Addendum: Regularization and variable selection via the elastic net

  2005cited by this paper
• 1-norm Support Vector Machines

  2003cited by this paper
• Stability and Generalization

  2002cited by this paper
• Feature selection algorithms: a survey and experimental evaluation

  2002cited by this paper
• A gene-expression signature as a predictor of survival in breast cancer.

  2002cited by this paper
• Regression Shrinkage and Selection via the Lasso

  1996influential reference
• On the Douglas—Rachford splitting method and the proximal point algorithm for maximal monotone operators

  1992cited by this paper

• A systematic review of networks for prognostic prediction of health outcomes and diagnostic prediction of health conditions within Electronic Health Records

  2024cites this paper
• Memetic Evolution of Training Sets with Adaptive Radial Basis Kernels for Support Vector Machines

  2021cites this paper
• Graph-Representation of Patient Data: a Systematic Literature Review

  2020cites this paper
• Stability of Indexed Microarray and Text Data

  2017cites this paper