Konferenční příspěvek

Somol P., Novovičová J., Grim J., Pudil P.

ICPR 2008 Proceedings (Int. Conf. on Pattern Recognition)

IEEE Computer Society

Tampa, Florida

2008

RNDr. Petr Somol, Ph.D. (ÚTIA - Oddělení rozpoznávání obrazů)

feature selection, subset search, sequential search, oscilla

We introduce a new feature selection method suitable for non-monotonic criteria, i.e., for Wrapper-based feature selection. Inspired by Oscillating Search, the Dynamic Oscillating Search: (i) is deterministic, (ii) optimizes subset size, (iii) has built-in preference of smaller subsets, (iv) has higher optimization performance than other sequential methods. We show that the new algorithm is capable of over-performing older methods not only in criterion maximization ability but in some cases also in obtaining subsets that generalize better.