19.11 Neural networks and support vector methods

On neural networks, some theoretical background can be found in Kecman (2001), Vidyasagar (2002) and De Wilde (1997). The application of neural networks in dynamic modelling, estimation and control is treated in Hovakimyan et al. (2000), Norgaard et al. (2003) and Pham and Liu (1995).

Considerable claims are being made for the achievements of new rivals to neural networks in the form of support vector methods, kernel methods and adaptive logic networks. See sample references Kecman (2001), Lee and Verri (2002), Cristianini and Shawe-Taylor (2000) or Sch lkopf et al. (1999).