Collection of projects completed during my Master's course in Deep Learning. All the projects use the Pytorch library to implement different Neural Network architectures. The following projects are present, both with python scripts and an explanatory Jupyter notebook:
- Feed-ForwardMNIST: Simple approach to the handwritten digit recognition trough a feed forward neural network;
- LSTMtextgenerator: An implementation of a text generator trough Long Short Term Memory Recurrent Neural Network;
- ConvAutoencodersMNIST: Implementation of a convolutional neural network which uses autoencoders to encode and then reconstruct digit from the MNIST dataset;
- ReinforcementBasic: Basic exercise of reinforcement learning where an agent needs to reach a certain goal in the environment.