Optimizing Shallow Networks for Binary Classification
This work addresses the problem of developing alternative training methods for neural network classification, though it appears incremental as it builds on existing mathematical analyses without demonstrating broad SOTA impact.
The authors introduced a new family of optimization problems for training neural networks in binary classification, which differs from existing approaches and enables novel algorithms with simple implementation and stable convergence.
Data driven classification that relies on neural networks is based on optimization criteria that involve some form of distance between the output of the network and the desired label. Using the same mathematical analysis, for a multitude of such measures, we can show that their optimum solution matches the ideal likelihood ratio test classifier. In this work we introduce a different family of optimization problems which is not covered by the existing approaches and, therefore, opens possibilities for new training algorithms for neural network based classification. We give examples that lead to algorithms that are simple in implementation, exhibit stable convergence characteristics and are antagonistic to the most popular existing techniques.