I have trie coding myself while following your nice video!

If you have time, please check this out.

I have added Relu and sigmoid activation function as well.

https://github.com/Rowing0914/neural_network_cplusplus

Best,

Norio

Great tutorial for a beginner like me.

I have done the implementation in Java.

Sharing below for anyone who might be interested.

https://github.com/sandeepyadavkec/BasicNeuralNetJava

Thanks again.

]]>I’m trying to get the network to work on handwriting training examples. I’ve vectorised images that are 28×28, so I have 784 inputs into the network and I want to classify images as 0, 1 … 9.

I get the network to run over a large number of training examples but for each pattern it always outputs 1. I noticed in one of your comments you said to scale down the initial weights which I have done but still the same output.

Any ideas? Thanks

]]>where did you get the formula with smoothing factor from? Thanks! ]]>

Thanks in advance!

]]>Thanks in advance!

]]>Best wishes! ]]>

Hello Dave. Thank you very much for the prompt reply. Your explanation was really helpful. There is still very much to learn for me but you have given me a very easy start. Keep the good work going! Once again thank you. ]]>