Future Work

The next step in this project is feed forward neural networks implementation.

In fact, we have developped all the tools required to train a neural net, we only need to specify the weights of the model as input nodes and implement new blocks : convolution and several other layers.

In terms of applications, we think about medical image processing : using convolutional neural networks, we can detect and segment some regions on a medical X-ray. This tool can help the doctors to better cure the patients, as it will have access to an unbiased ML algorithm to perform segmentations.