Use Case of Counterfactual Examples: Data Augmentation

Abstract

Counterfactual explanations are gaining popularity as a way of explaining machine learning models. Counterfactual examples are generally created to interpret the decision of a model. In this case, if a model makes a certain decision for an instance, the counterfactual examples of that instance reverse the decision of the model. The counterfactual examples can be created by craftily changing particular feature values of the instance. In this work, we explore other potential application areas of utilizing counterfactual examples other than model explanation. We are particularly interested in exploring whether counterfactual examples can be a good candidate for data augmentation. At the same time, we look for ways of validating the generated counterfactual examples.