A Modified Constructive Neural Networks and Its Application for Large-scale Data Mining

The constructive neural networks based on the covering algorithms is suitable for large-scale data mining because it can be local processing and has little computational complexity, however, the local processing lows classification precision. In this paper, covering algorithms is firstly extended to kernel covering algorithms and we secondly construct a kind of finite mixture probabilistic model based on kernel covering algorithms according to the probability meaning of Gaussian function and finally introduce the global optimizing computation by "maximum likelihood theory", realize the global optimization problem of the covering algorithms so as to expand the range of application of covering algorithms and improve its precision. The results of the experiment are given as an example to illustrate the validation of the method.