Home

Calendar

Filter

Search

Mechanical Engineering Graduate Seminar: Tong Wang

Apr 18, 2019

03:30 PM - 04:20 PM

Seamans Center, Room 3505

103 South Capitol Street, Iowa City, IA 52240

Save to My Events

Title: Hybrid Predictive Model: When an Interpretable Model Collaborates with a Black-box Model

Presented by: Tong Wang, Assistant Professor

Management Sciences, Tippie College of Business

 

Abstract:

Interpretable machine learning has received increasing interest in recent years, especially in domains where humans are involved in the decision-making process. However, the possible loss of the task performance for gaining interpretability is often inevitable, especially for large datasets or complicated tasks. This performance downgrade puts practitioners in a dilemma of choosing between a top-performing black-box model and an interpretable model with unsatisfying task performance. In this work, we propose a novel framework for building a Hybrid Predictive Model (HPM) that integrates an interpretable model with any black-box model to introduce interpretability in the decision-making process at no or low cost of the predictive accuracy. The interpretable model substitutes the black-box model on a subset of data where the black-box model is overkill or nearly overkill. We design a principled objective function that considers predictive accuracy, model interpretability, and model transparency, which is the percentage of data processed by the interpretable model. This framework brings together the advantages of the high predictive performance of black-box models and the high interpretability of interpretable models. Under this framework, we propose a hybrid rule set model that uses association rules as the interpretable substitute and design customized training algorithms with theoretically grounded bounds to reduce computation. We test the hybrid predictive models on structured datasets and text data.  In these experiments, the interpretable models collaborate with state-of-the-art black-box models including ensemble models and neural networks. We propose to use two efficient frontiers to characterize the trade-off between transparency and predictive performance and the trade-off between transparency and the average number of features used to make a prediction. Results show that hybrid models are able to obtain transparency at no or low cost of predictive performance using significantly fewer features.

Individuals with disabilities are encouraged to attend all University of Iowa–sponsored events. If you are a person with a disability who requires a reasonable accommodation in order to participate in this program, please contact in advance at