Topological Interpretation of Deep-Learning Models

Workshop:CAFCW21: Computational Approaches for Cancer Workshop 2021

Authors: Adam Spannaus, Shang Gao, and Noah Schaefferkoetter (Oak Ridge National Laboratory (ORNL)); Lynne Penberthy (National Cancer Institute (NCI)); Jennifer Doherty (Utah Cancer Registry); Eric Durbin (Kentucky Cancer Registry); Steven Schwartz (Seattle-Puget Sound Registry); Antoinette Stroup (New Jersey State Cancer Registry); Xiao-Cheng Wu (Louisiana Tumor Registry); Charles Wiggins (New Mexico Tumor Registry); and Georgia Tourassi (Oak Ridge National Laboratory (ORNL))

Abstract: Developing trust in the predictions made from an AI-based algorithm is a tantamount concern, especially in systems such as threat detection or medical diagnosis where outcomes may have tragic consequences. This work presents a topologically informed methodology for inferring prominent features in a deep-learning classification model trained on clinical text. Creating a graph of the model's prediction space, we cluster the inputs into the vertices of the graph, and extract subgraphs demonstrating high-predictive accuracy for a given label. These nodes contain a wealth of information about features that the deep-learning model has recognized as important. We demonstrate fidelity of these inferred features to the original model by appending these terms to incorrectly classified documents, and showing an increase in classification accuracy on these augmented documents. This work demonstrates that we may gain actionable insights for subsequent models by improving erroneous classifications and providing a path forward for continuous improvement.

Website:

Back to CAFCW21: Computational Approaches for Cancer Workshop 2021 Archive Listing

Back to Full Workshop Archive Listing