Optimal Sparse Survival Trees.

Interpretability is crucial for doctors, hospitals, pharmaceutical companies and biotechnology corporations to analyze and make decisions for high stakes problems that involve human health. Tree-based methods have been widely adopted for due to their appealing interpretablility and their ability to capture complex relationships. However, most existing methods to produce survival trees rely on heuristic (or greedy) algorithms, which risk producing sub-optimal models.

Optimal Sparse Regression Trees.

Regression trees are one of the oldest forms of AI models, and their predictions can be made without a calculator, which makes them broadly useful, particularly for high-stakes applications. Within the large literature on regression trees, there has been little effort towards full provable optimization, mainly due to the computational hardness of the problem. This work proposes a dynamic-programming-with-bounds approach to the construction of provably-optimal sparse regression trees.

Are teachers meeting students' needs in untracked science classrooms? Evidence based on a causal inferential approach.

Tracking has been criticized for relegating disadvantaged students to lower track courses in which students encounter a greater lack of instructional support. While an end to tracks through detracking is a possible solution, there are concerns that detracking will create more heterogeneous classrooms, making it harder for teachers to provide adequate support to their students. Using the 2015 PISA dataset, this study conducts a causal inferential analysis to understand the differences in student perceptions of teaching in tracked and untracked environments.

Exploring and Interacting with the Set of Good Sparse Generalized Additive Models.

In real applications, interaction between machine learning models and domain experts is critical; however, the classical machine learning paradigm that usually produces only a single model does not facilitate such interaction. Approximating and exploring the Rashomon set, i.e.