An Application of Component-Wise Iterative Optimization to Feed-Forward Neural Networks

摘要: Component-wise Iterative Optimization (CIO) is a method of dealing with large data in the OLAP applications, which can be treated as enhancement traditional batch version methods such least squares. The salient feature to process transactions one by one, optimizes estimates iteratively for each parameter over given objective function, and update models on fly. A new learning algorithm proposed when applying CIO feed-forward neural networks single hidden layer. It incorporates internal structure layer closed-form expressions weights between output Its optimally computational property natural consequence inherited from also demonstrated an illustrative example.