Approximating Viability Kernels With Support Vector Machines
Authors:
- 15 May 2007
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Automatic Control
- Vol. 52 (5) , 933-937
Abstract
We propose an algorithm which performs a progressive approximation of a viability kernel, iteratively using a classification method. We establish the mathematical conditions that the classification method should fulfil to guarantee the convergence to the actual viability kernel. We study more particularly the use of support vector machines (SVMs) as classification techniques. We show that they make possible to use gradient optimisation techniques to find a viable control at each time step, and over several time steps. This allows us to avoid the exponential growth of the computing time with the dimension of the control space. It also provides simple and efficient control procedures. We illustrate the method with some examples inspired from ecologyKeywords
This publication has 9 references indexed in Scilit:
- Autonomous reactive control for simulated humanoids2004
- VIABILITY MODEL OF TROPHIC INTERACTIONS IN MARINE ECOSYSTEMSNatural Resource Modeling, 2004
- Approximate safety enforcement using computed viability envelopes2004
- Making ecosystem models viableBulletin of Mathematical Biology, 2003
- A viability analysis for a bio-economic modelEcological Economics, 2001
- Programming and control of robots by means of differential algebraic inequalitiesIEEE Transactions on Robotics and Automation, 2000
- An Introduction to Support Vector Machines and Other Kernel-based Learning Methods2000
- Neural Networks and Qualitative Physics1996
- Approximation of the viability kernelApplied Mathematics & Optimization, 1994