Hydrocarbon prediction using neural networks of structural risk minimization

摘要

In this paper, the neural network of structural risk minimization is proposed based on the structural risk minimization principle of Statistical Learning Theory. This method automatically expands the capacity of neural networks, and then the structural design of neural networks achieved. It can improve the accuracy of networks training and the generalization of the neural networks when the training sample is limited. The practical applications showed that the neural network of structural risk minimization obtains better results in hydrocarbon prediction.