Artificial Neural Network library in C. Made in 10 days as a submission to Code Guessing #57. The files I've submitted can be viewed in code_guessing branch.
The dataset I've used to train my network is a modified version of this dataset.
This library implements dense and dropout(*) layers, SGD algorithm, four activation functions, mean squared error and categorical cross entropy loss functions.
* The dropout layer doesn't really work, so I've temporarily added a parameter to dense layers. I might update it in the future
I wanted to understand how ANNs really worked by implementing something like this. Code Guessing #57 provided a good opportunity for me to do so.
I wanted it to be more challenging.