This repository contains all the main projects I have been assesed in Udacity's Nanodegree program for Deep Learning. This program has taught me how to implement the main types of neural networks (FFNN,CNN,RNN,Auto Encoders,GANs,DCGANs, etc) in Pytorch along with relevant concepts in the fields of Data Science and Machine Learning such as attention, batching, data pre-processing and so on.
The projects of this repository are:
- A Feed Forward Neural Network (FFNN) for predicting bike sharing patterns and prices.
- Convolutional Neural Network (CNN) for image recognition in a program to classify dog breeds and find similarities between human faces and dog breeds.
- Recurrent Neural Networks (RNN) for sequence processing to generate a TV script based on the plot of a TV Series.
- A Deep Convolutional Generative Adversial Network (DCGAN) to generate human faces.
- Deploying a Recurrent Neural Network for movie sentiment analysis on the web using Amazon Web Services (AWS) SageMaker, Lambda, API Gateaway and Cloudfront.
In each project's folder, you will find a descriptive README where it will describe the purpose of the project and its implementation. Instructions on how to run it will also be provided.
MIT License
Copyright (c) 2020 Mateo Randulfe