What are the trade-offs between interpretability and predictive performance in a high-stakes clinical setting for BigHat Biosciences?
