The Bayesian Case Model: A Generative Approach for Case-Based Reasoning and Prototype Classification

• Main Authors: Kim, Been (Contributor), Rudin, Cynthia (Contributor), Shah, Julie A. (Contributor)
• Other Authors: Massachusetts Institute of Technology. Department of Aeronautics and Astronautics (Contributor), Sloan School of Management (Contributor)
• Format: Article
• Language: English
• Published: Neural Information Processing Systems Foundation, Inc., 2014-11-26T14:48:31Z.
• Subjects: Article
• Online Access: Get fulltext

 We present the Bayesian Case Model (BCM), a general framework for Bayesian case-based reasoning (CBR) and prototype classification and clustering. BCM brings the intuitive power of CBR to a Bayesian generative framework. The BCM learns prototypes, the "quintessential" observations that best represent clusters in a dataset, by performing joint inference on cluster labels, prototypes and important features. Simultaneously, BCM pursues sparsity by learning subspaces, the sets of features that play important roles in the characterization of the prototypes. The prototype and subspace representation provides quantitative benefits in interpretability while preserving classification accuracy. Human subject experiments verify statistically significant improvements to participants' understanding when using explanations produced by BCM, compared to those given by prior art.