Institute of Information Theory and Automation

You are here

Publication details

Efficient Feature Subset Selection and Subset Size Optimization

Monography Chapter

Somol Petr, Novovičová Jana, Pudil Pavel

serial: Pattern Recognition, Recent Advances, p. 75-98 , Eds: Herout A.

research: CEZ:AV0Z10750506

project(s): 1M0572, GA MŠk, 2C06019, GA MŠk, GA102/08/0593, GA ČR, GA102/07/1594, GA ČR

keywords: dimensionality reduction, pattern recognition, machine learning, feature selection, optimization, subset search, classification

preview: Download

abstract (eng):

A broad class of decision-making problems can be solved by learning approach. This can be a feasible alternative when neither an analytical solution exists nor the mathematical model can be constructed. In these cases the required knowledge can be gained from the past data which form the so-called learning or training set. Then the formal apparatus of statistical pattern recognition can be used to learn the decision-making. The first and essential step of statistical pattern recognition is to solve the problem of feature selection (FS) or more generally dimensionality reduction (DR). The chapter summarizes the state of art in feature selection, addressing key topics including: FS categorization, FS criteria, FS search strategies, FS stability.


2012-12-21 16:10