Chapter 7 Regularization Theory 《神经网络与 及机器学习》 .pptVIP

Chapter 7 Regularization Theory;Figure 7.1 Illustration of the mapping of (input) domain X onto (output) range Y.;Figure 7.2 Regularization network.;Figure 7.3 Radial-basis-function network of reduced complexity.;Figure 7.4 Model of the semisupervised learning process.;Table 7.1;Figure 7.7 Lapacian RLS classification of the double-moon of Fig. 1.8 with distance d = –1 and two labeled data points per class that are denoted by the markers ? and ?. Intrinsic regularization parameter λI = 0.0001.;Figure 7.8 Laplacian RLS classification of the double-moon of Fig. 1.8 with distance d = –1 and two labeled data points that are denoted by the markers ? and ?. The intrinsic regularization parameter λI = 0.1.;Figure 7.9 USPS data classification using (a) RLS algorithm and (b) Laplician RLS algorithm. (Reproduced with the permission of Dr.Vikas Sindhwani.)