SCAWI: an algorithm for weight initialization of a sigmoidal neural network
暂无分享,去创建一个
A weight initialization procedure, which strongly improves the performance of the BP algorithm is suggested. An optimum range, into which the weights are randomly chosen, is shown to exist in order to minimize the time needed to reach the minimum of the cost function.
[1] Sandro Ridella,et al. An optimum weights initialization for improving scaling relationships in BP learning , 1991 .
[2] Sandro Ridella,et al. Statistically controlled activation weight initialization (SCAWI) , 1992, IEEE Trans. Neural Networks.
[3] Christian Lebiere,et al. The Cascade-Correlation Learning Architecture , 1989, NIPS.