%0 Journal Article %1 venkatraman:2004:CIM %A Venkatraman, Vishwesh %A Dalby, Andrew Rowland %A Yang, Zheng Rong %D 2004 %J Journal of Chemical Information and Modeling %K algorithms, genetic programming %N 5 %P 1686--1692 %R doi:10.1021/ci049933v %T Evaluation of Mutual Information and Genetic Programming for Feature Selection in QSAR %V 44 %X Feature selection is a key step in Quantitative Structure Activity Relationship (QSAR) analysis. Chance correlations and multicollinearity are two major problems often encountered when attempting to find generalised QSAR models for use in drug design. Optimal QSAR models require an objective variable relevance analysis step for producing robust classifiers with low complexity and good predictive accuracy. Genetic algorithms coupled with information theoretic approaches such as mutual information have been used to find near-optimal solutions to such multicriteria optimisation problems. In this paper, we describe a novel approach for analyzing QSAR data based on these methods. Our experiments with the Thrombin dataset, previously studied as part of the KDD (Knowledge Discovery and Data Mining) Cup 2001 demonstrate the feasibility of this approach. It has been found that it is important to take into account the data distribution, the rule "interestingness", and the need to look at more invariant and monotonic measures of feature selection.

@article{venkatraman:2004:CIM, abstract = {Feature selection is a key step in Quantitative Structure Activity Relationship (QSAR) analysis. Chance correlations and multicollinearity are two major problems often encountered when attempting to find generalised QSAR models for use in drug design. Optimal QSAR models require an objective variable relevance analysis step for producing robust classifiers with low complexity and good predictive accuracy. Genetic algorithms coupled with information theoretic approaches such as mutual information have been used to find near-optimal solutions to such multicriteria optimisation problems. In this paper, we describe a novel approach for analyzing QSAR data based on these methods. Our experiments with the Thrombin dataset, previously studied as part of the KDD (Knowledge Discovery and Data Mining) Cup 2001 demonstrate the feasibility of this approach. It has been found that it is important to take into account the data distribution, the rule {"}interestingness{"}, and the need to look at more invariant and monotonic measures of feature selection.}, added-at = {2008-06-19T17:46:40.000+0200}, author = {Venkatraman, Vishwesh and Dalby, Andrew Rowland and Yang, Zheng Rong}, biburl = {https://www.bibsonomy.org/bibtex/25ed323ebbeff9cf59306434dde0d25b6/brazovayeye}, doi = {doi:10.1021/ci049933v}, interhash = {08424be7b38c7613e128df8537bbb182}, intrahash = {5ed323ebbeff9cf59306434dde0d25b6}, journal = {Journal of Chemical Information and Modeling}, keywords = {algorithms, genetic programming}, notes = {http://pubs.acs.org/journals/jcisd8/index.html American Chemical Society S0095-2338(04)09933-0 School of Biological Sciences, University of Exeter, Exeter EX4 4QF, Great Britain and School of Engineering and Computer Science, University of Exeter, Exeter EX4 4QF, Great Britain PMID: 15446827}, number = 5, pages = {1686--1692}, timestamp = {2008-06-19T17:53:39.000+0200}, title = {Evaluation of Mutual Information and Genetic Programming for Feature Selection in {QSAR}}, volume = 44, year = 2004 }