Support Vector Machines for Predictive Modeling in Heterogeneous Catalysis: A Comprehensive Introduction and Overfitting Investigation Based on Two Real Applications
- 13 May 2006
- journal article
- research article
- Published by American Chemical Society (ACS) in Journal of Combinatorial Chemistry
- Vol. 8 (4) , 583-596
Abstract
This works provides an introduction to support vector machines (SVMs) for predictive modeling in heterogeneous catalysis, describing step by step the methodology with a highlighting of the points which make such technique an attractive approach. We first investigate linear SVMs, working in detail through a simple example based on experimental data derived from a study aiming at optimizing olefin epoxidation catalysts applying high-throughput experimentation. This case study has been chosen to underline SVM features in a visual manner because of the few catalytic variables investigated. It is shown how SVMs transform original data into another representation space of higher dimensionality. The concepts of Vapnik−Chervonenkis dimension and structural risk minimization are introduced. The SVM methodology is evaluated with a second catalytic application, that is, light paraffin isomerization. Finally, we discuss why SVMs is a strategic method, as compared to other machine learning techniques, such as neural networks or induction trees, and why emphasis is put on the problem of overfitting.Keywords
This publication has 25 references indexed in Scilit:
- Statistical practice in high-throughput screening data analysisNature Biotechnology, 2006
- Ligand Descriptor Analysis in Nickel‐Catalysed Hydrocyanation: A Combined Experimental and Theoretical StudyAdvanced Synthesis & Catalysis, 2005
- Development of catalyst libraries for total oxidation of methaneApplied Catalysis A: General, 2005
- A statistical approach to class separabilityApplied Stochastic Models in Business and Industry, 2005
- Design of Discovery Libraries for Solids Based on QSAR ModelsQSAR & Combinatorial Science, 2005
- Improving process operations using support vector machines and decision treesAIChE Journal, 2005
- Design efficient support vector machine for fast classificationPattern Recognition, 2005
- Fundamental insights into the oxidative dehydrogenation of ethane to ethylene over catalytic materials discovered by an evolutionary approachCatalysis Today, 2003
- Neural Networks and the Bias/Variance DilemmaNeural Computation, 1992
- Asymptotics for and against cross-validationBiometrika, 1977