Optimal Decision Trees (ODT)
This package was built with my team based on the algorithm described in the paper “Precision oncology, a review to assess interpretability in several explainable methods.” ODT builds an optimal decision tree by solving an optimization problem at each split, assigning treatments that optimize a chosen sensitivity measure while respecting group size constraints. The method leverages clinical/omics and drug sensitivity data to recommend robust, individualized therapies.
- Focus: interpretable, optimization-driven treatment recommendations
- Stack: R, optimization, clinical/omics data integration