Computed Decision Weights and a New Learning Algorithm for Neural Classifiers
This addresses the challenge of weight computation in neural networks for researchers and practitioners, but it appears incremental as it builds on existing optimization methods.
The paper tackles the problem of computing decision layer weights for neural classifiers instead of training them, achieving this through a specific loss function and constrained optimization, which results in a new learning process for pre-decision weights that is both simple and effective.
In this paper we consider the possibility of computing rather than training the decision layer weights of a neural classifier. Such a possibility arises in two way, from making an appropriate choice of loss function and by solving a problem of constrained optimization. The latter formulation leads to a promising new learning process for pre-decision weights with both simplicity and efficacy.