Connections Between Optimisation-Based Regressor Selection and Analysis of Variance
2006 (English)In: Proceedings of Reglermöte 2006, 2006Conference paper (Other academic)
Earlier contributions have shown that Analysis of Variance (ANOVA) can be successfully used for finding good regressors for nonlinear models in a nonlinear black-box system identification context. In this paper, it is shown that the ANOVA problem can be recast as an optimisation problem. A modified, convex version of the ANOVA optimisation problem is then proposed, and it turns out that it is closely related to the nn-garrote method, respectively. In the case of balanced data, it is also shown that the method has a nice orthogonality property in the sense that different groups of parameters can be computed independently.
Place, publisher, year, edition, pages
Regressor selection, ANOVA, 1-norm, Lasso, Nn-garrote
Engineering and Technology Control Engineering
IdentifiersURN: urn:nbn:se:liu:diva-36931Local ID: 33078OAI: oai:DiVA.org:liu-36931DiVA: diva2:257780
Reglermöte 2006, Stockholm, Sweden, May, 2006