VAC Colloquium: Sascha Marton "Learning Axis-Aligned Decision Trees with Gradient Descent"

Sascha Marton, scientific researcher at the University of Mannheim's Institute for Enterprise Systems (InES), will give a presentation on the topic

"Learning Axis-Aligned Decision Trees with Gradient Descent"

as part of the VAC Colloquium.

Afterwards we look forward to discussions while enjoying coffee and cookies.

The event is open to anyone interested.

The presentation will be held via Zoom, you can attend in person or by clicking the link:

Decision Trees (DTs) are commonly used for many machine learning tasks due to their high degree of interpretability. However, learning a DT from data is a difficult optimization problem, as it is non-convex and non-differentiable. Therefore, common approaches learn DTs using a greedy growth algorithm that minimizes the impurity locally at each internal node. Unfortunately, this greedy procedure can lead to inaccurate trees. In this talk, I present a novel approach for learning hard, axis-aligned DTs with gradient descent. The proposed method uses backpropagation with a straight-through operator on a dense DT representation, to jointly optimize all tree parameters. The approach outperforms existing methods on binary classification benchmarks and achieves competitive results for multi-class tasks.


Back to Eventlist