Art design teaching assessment based on improved fuzzy robust wavelet RVM algorithm
An improved fuzzy robust wavelet relevance vector machine can be proposed and applied to art design teaching assessment in this study. Relevance vector machine is a Bayesian extension of the SVM, which can solve the over-fitting of artificial neural networks, and triangular fuzzy, robust loss function and Morlet wavelet function can be used to construct a novel relevance vector machine. In order to testify that the improved fuzzy wavelet relevance vector machine has higher prediction performance than traditional relevance vector machine, the testing case is applied to show the superiority of the improved fuzzy wavelet relevance vector machine compared with traditional relevance vector machine. The comparison of mean assessment error between improved fuzzy robust wavelet RVM and classical RVM is given in the experiment. It can be seen that mean assessment error of improved fuzzy robust wavelet RVM is lower than classical RVM. Thus, we can conclude that mean assessment ability of improved fuzzy robust wavelet RVM is better than that of classical RVM.