P
Petr Somol
Researcher at Academy of Sciences of the Czech Republic
Publications - 71
Citations - 1872
Petr Somol is an academic researcher from Academy of Sciences of the Czech Republic. The author has contributed to research in topics: Feature selection & Dimensionality reduction. The author has an hindex of 19, co-authored 68 publications receiving 1779 citations. Previous affiliations of Petr Somol include University of Economics, Prague & Charles University in Prague.
Papers
More filters
Journal ArticleDOI
Adaptive floating search methods in feature selection
Petr Somol,Petr Somol,Petr Somol,Pavel Pudil,Pavel Pudil,Jana Novovičová,Jana Novovičová,P. Paclík,P. Paclík +8 more
TL;DR: A new suboptimal search strategy for feature selection that represents a more sophisticated version of “classical” floating search algorithms and facilitates finding a solution even closer to the optimal one.
Journal ArticleDOI
Fast branch & bound algorithms for optimal feature selection
TL;DR: A novel search principle for optimal feature subset selection using the branch & bound method using a simple mechanism for predicting criterion values is introduced and two implementations of the proposed prediction mechanism are proposed that are suitable for use with nonrecursive and recursive criterion forms.
Journal ArticleDOI
Road sign classification using Laplace kernel classifier
TL;DR: A new kernel rule has been developed for road sign classification using the Laplace probability density and an Expectation–Maximization algorithm is used to maximize the pseudo-likelihood function.
Journal ArticleDOI
Evaluating Stability and Comparing Output of Feature Selectors that Optimize Feature Subset Cardinality
Petr Somol,Jana Novovičová +1 more
TL;DR: This work investigates the problem of evaluating the stability of feature selection processes yielding subsets of varying size and introduces several novel feature selection stability measures and adjusts some existing measures in a unifying framework that offers broad insight into the stability problem.
Book ChapterDOI
Feature selection based on mutual correlation
TL;DR: A novel filter feature selection method based on mutual correlation is proposed that is assessed by using the selected features to the Bayes classifier and a trade off between the classification accuracy and the feature set dimensionality is demonstrated.