The search space for the feature selection problem in decision tree learning is the lattice of subsets of the available features. We provide an exact enumeration procedure of the subsets that lead to all and only the distinct decision trees. The procedure can be adopted to prune the search space of complete and heuristics search methods in wrapper models for feature selection. Based on this, we design a computational optimization of the sequential backward elimination heuristics with a performance improvement of up to 100X.
Salvatore Ruggieri (Università di Pisa)
Related Events (a corresponding poster, oral, or spotlight)
2017 Talk: Enumerating Distinct Decision Trees »
Mon Aug 7th 12:30 -- 12:48 AM Room C4.6 & C4.7