Using a support vector machine and sampling to classify compounds as potential transdermal enhancers

  • Authors:

  • Alpa Shah;Gary P. Moss;Yi Sun;Rod Adams;Neil Davey;Simon Wilkinson

  • Affiliations:

  • Science and technology research school, University of Hertfordshire, United Kingdom;School of Pharmacy, Keele University, United Kingdom;Science and technology research school, University of Hertfordshire, United Kingdom;Science and technology research school, University of Hertfordshire, United Kingdom;Science and technology research school, University of Hertfordshire, United Kingdom;Medical Toxicology Centre, Wolfson Unit, Medical School, University of Newcastle-upon-Tyne, UK

  • Venue:
  • ICANN'12 Proceedings of the 22nd international conference on Artificial Neural Networks and Machine Learning - Volume Part II
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

Distinguishing good chemical enhancers of percutaneous absorption from poor enhancers is a difficult problem. Previously, discriminant analysis and other machine learning methods have been applied to this problem. Results showed that the ordinary SVM provided the best result. In this work, we apply both SVM with different cost errors and sampling methods to improve the accuracy of classification. We show that a good classification is possible.