This repository contains the implementations for the course “Topics in Deep Learning” @ CMU (10-707). All the implementations are in python.
Implement a multi-layer neural network with different activation funcations and mini-batched data with batch normalisation.
Implement a Restricted Boltzmann Machine and an Autoencoder and use them as pre-trainers for the input to the MLP.
Build a 4-gram Neural Language Model and interpret the word vectors learnt from it.