This dataset contains replication data for the paper Regularized Feature Selection Landscapes: An Empirical Study of Multimodality. It contains accuracy tables of well-known classification datasets from The UCI Machine Learning Repository.
These tables comprise the accuracy for all feature subsets, i.e., all column combinations, obtained by a decision tree classifier, with different levels of regularization.
The decision tree classifier is provided by the BetaML.jl package for the Julia programming language, and uses the default parameters: max_depth=0 (no limits) min_gain=0.0, min_records=2, max_features=0 (consider all features), splitting_criterion=BetaML.Utils.gini, (Gini impurity index), and rng = Random.GLOBAL_RNG (no set seed for random number generation).
Julia, 1.9.3