Deep Learning ND uploaded here for public viewing of the completed projects and the skills gained.
Project 1: Dog Breed Classifier
Used Convolution neural Networks to predict the breed of the dog. Also used Haar cascade classifier to detect images and transfer learning from Resnet-50. If an image of human is passed, it will tell the closest breed of the dog the said human resembles to :p
Project 2: TV Script Generation
Used MultiRNN and BasicLSTM cells to generate TV Script for Simpsoms show.
Project 3: Face Generation using Generative Adversial Networks
A Discriminator and Generator implemented on the MNIST and CelebA datasets.
Project 4: Training a Quadcopter to how to fly using Reinforcement Learning
Please checkout RL-Quadcopter 2-7.html, tasks.py and agents folder. Used Actor-Critic method here. I implemented the project to make sure that Quadcopter remains stationary and stabvle at one point.