Enclosing Machine Learning

摘要: As known to us, the cognition process is instinct learning ability of human being. This perhaps one most complex behaviors. It a highly efficient and intelligent information processing process. For natural world, humans always transfer feature brain through their perception, then processes remembers it. Due invention computers, scientists are now working toward improving its artificial intelligence, they hope that day computer could have “brain” as does. However, it still long way for us go in order let truly “think” by itself. Currently, intelligence an important active research topic. imitates using idea function equivalence. Traditionally, neural computing networks families majority parts direction (Haykin, 1994). By imitating mechanism human-brain neuron, built theory following experimental such perception neurons spiking (Gerstner & Kistler, 2002) understand neurons. Neural-computing (NN) (Bishop, 1995) made great achievements various aspects. Recently, statistical support vector machines (SVM) (Vapnik, drawn extensive attention shown better performances areas (Li, Wei Liu, 2004) than NN, which implies can also be via advanced theory. Nowadays, these two methods tend merge under framework. BACKGROUND