A fast variable selection for nonnegative garrote-based artificial neural network

This paper proposes a new early stopping algorithm for improving processing time of nonnegative garrote (NNG)-artificial neural network (ANN) variable selection method which has been used for reducing model complexity in soft sensor application. The performance of the proposed method is compared with conventional NNG-ANN variable selection algorithm. The experiments from two simulation cases are used for demonstrating the performance of the method and the experimental results show better processing times, while maintaining the capability of correctly selecting the set of significant input variables.