CODEWITHSUNDEEP
neural-networkbackpropagationfeed-forward
drinu16
drinu16

Reputation: 795

Training a FeedForward Neural Network

I have implemented a Back propagation Neural Network, now I would like to implement a Feed Forward Neural Network to compare their accuracy.

My question is, what learning methods does Feed Forward has (apart from back propagation), because every article mentions back propagation as the learning method.

Upvotes: 1

Views: 534

Answers (1)

jorgenkg
jorgenkg

Reputation: 4275

A possible misconception

A "feed forward neural network" is a neural network without recurrent connections. The name is a description of how the input signal are propagated throughout the network structure. Thus, you've already implemented a feed forward network.

The backpropagation algorithm is a training (or a weight adjustment) algorithm that can be used to teach a feed forward neural network how to classify a dataset.

Other learning strategies

Genetic algorithm

Another popular learning strategy is using the genetic algorithm. This is not as computationally effective, but it does not suffer from becoming stuck at local optima. This seems to be a great introduction.

Hebbian learning

This is an old fashioned AI learning algorithm, which has gained some ground again. You can read about Hebbian learning here.

Upvotes: 1

Related Questions