Counterfactual Explanations for Trial Outcome Prediction

Sunstella 2023 Summer Research Camp Submission3 Authors

Keywords: Counterfactual, Deep Learning, Bert, TextCNN

Abstract: This study introduces a novel framework for enhancing the interpretability of machine learning models in trial prediction. By leveraging counterfactual explanations, the framework highlights the key factors influencing predictions and provides actionable insights for medical practitioners and researchers. Our objectives are to develop an accurate deep learning model for clinical trial outcome prediction and generate interpretable counterfactual explanations. Through evaluation, we aim to bridge the gap between machine learning predictions and real-world decision-making, empowering medical professionals with a valuable tool for improving clinical trial outcomes.

Submission Number: 3