240 likes | 379 Views
Filter methods . T statistic Information Distance Correlation Separability …. FSS Algorithms. Exponential Exhaustiva search Branch & Bound Beam search Sequential SFS and SBS Plus-l / Minus-r Bidirectional Floating (exponential in worst case) Randomized Sequential + randomness
E N D
Filter methods • T statistic • Information • Distance • Correlation • Separability … Data Mining 2, 2006
FSS Algorithms • Exponential • Exhaustiva search • Branch & Bound • Beam search • Sequential • SFS and SBS • Plus-l / Minus-r • Bidirectional • Floating (exponential in worst case) • Randomized • Sequential + randomness • Genetic Data Mining 2, 2006
SFS performs best when the optimal subset has a small number of features When the search is near the empty set, a large number of states can be potentially evaluated Towards the full set, the region examined by SFS is narrower since most of the features have already been selected Data Mining 2, 2006
Example The optimal feature subset turns out to be {x1, x4}, because x4 provides the only information that x1 needs: discrimination between classes ω4 and ω5 Data Mining 2, 2006
SBS works best when the optimal feature subset has a large number of features, since it spends most of its time visiting large subsets The main limitation of SBS is its inability to reevaluate the usefulness of a feature after it has been discarded Data Mining 2, 2006
SFS is performed from the empty set SBS is performed from the full set Features selected by SFS are not removed by SBS Features removed by SBS are not selected by SFS Guarantee convergence Data Mining 2, 2006
Some backtracking ability Main limitation is that there is no theoretical way of predicting the optimal values of l and r Data Mining 2, 2006
Monotonicity in J: Data Mining 2, 2006
B A=J({1,2}) • The value of A is updated when a greater one is found in a leaf • Stop whenever every leaf has been purged or evaluated A >= B purge Data Mining 2, 2006
With a proper queue size, Beam Search can avoid getting trapped in local minimal by preserving solutions from varying regions in the search space The optimal is 2-3-4 (J=9), which is never explored Data Mining 2, 2006
The RELIEF algorithm (1) Data Mining 2, 2006
The RELIEF algorithm (2) Data Mining 2, 2006
Conclusions • Dificult and pervasive problem! • Lack of accepted and useful definitions for relevance, redundancy and irrelevance • Nesting problem • Abundance of algorithms and filters • Lack of proper comparative benchmarks • Obligatory step, usually well worth the pain Data Mining 2, 2006