On the Infeasibility of Training Neural Networks with Small Squared Errors

Vu, Van H.

–Neural Information Processing Systems 

We demonstrate that the problem of training neural networks with small (average) squared error is computationally intractable.

Neural Information Processing Systems

Dec-31-1998

Conferences    PDF

Add feedback

Duplicate Docs Excel Report

Title
On the Infeasibility of Training Neural Networks with Small Squared Errors
On the Infeasibility of Training Neural Networks with Small Squared Errors

Similar Docs  Excel Report  more

TitleSimilaritySource
None found