Problem with NN training (cause: min_gradient reached)

3 views (last 30 days)
Krzysztof
Krzysztof on 21 Aug 2013
Hi I've made FFB network with 3 layers, 3neurons on the 1st, 10 on the 2nd and 3 on the last one. Ive got a training set 17x8018matrix as an input and 3x8018 matrix as a output. default options for the rest, apart from weights' values which I set on 1 value. The training stops so quickly, after max 4loops. Cause: minimum gradient reached. How can I solve this problem?
Greets

Sign in to answer this question.

Answers (0)

Categories

Find more on Sequence and Numeric Feature Data Workflows in Help Center and File Exchange

Tags

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!