SAT-based Decision Tree Learning for Large Data Sets
André Schidler, Stefan Szeider
Abstract
Decision trees of low depth are beneficial for understanding and interpreting the data they represent. Unfortunately, finding a decision tree of lowest complexity (depth or size) that correctly represents given data is NP-hard. Hence known algorithms either (i) utilize heuristics that do not minimize the depth or (ii) are exact but scale only to small or medium-sized instances. We propose a new hybrid approach to decision tree learning, combining heuristic and exact methods in a novel way. More specifically, we employ SAT encodings repeatedly to local parts of a decision tree provided by a standard heuristic, leading to an overall reduction in complexity. This allows us to scale the power of exact SAT-based methods to comparatively very large data sets. We evaluate our new approach experimentally on a range of real-world instances that contain up to several thousand samples. In almost all cases, our method successfully decreases the complexity of the initial decision tree; often, the decrease is significant.