In neural nets, the back-propagation algorithm is 'based on' gradient descent, optimizing over a cost function $C(w,b)$ of the weights and biases of the network.

Conjugate gradient descent has way faster convergence than standard gradient descent (and stochastic gradient descent). Is there a similar back-propagation algorithm based off of conjugate gradient descent?


1 Answer 1



see the article Optimization in an error backpropagation neural network environment with a performance test on a spectral pattern classification problem


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.