Hello world notebook for getting started with deep learning. A step by step introduction to multi layer perceptron and how you can build a model in 5 simple steps using Keras and TensorFlow

If you are familiar with ML, you would have probably used sklearn. This notebooks will enable you to use the familiar sklearn library to tune a deep learning model, and also use K-Fold cross validation to check the stability of the model.

Moving beyond binary classification to multiclassification problem. Building a MLP model using a very simple tweak to binary classification model on Iris flower species dataset

Exploring shallow and deeper models with Keras and TensorFlow on Mines vs Rocks dataset and comparing the performances of different models

Moving from structured datasets to perceptual problems, building our first CNN model on the famous MNIST digit recognition dataset and submitting on Kaggle to visualize the performance. We will also compare GBM based model - XGBoost against the deep learning model

Image augmentation is a technique to artificially increase the size of data to make the model generalize well over varying scenarios. We will explore different kinds of augmentations that can be done with Keras ImageDataGenerator module

A simple implementation LSTM for solving NLP problem of text classifcation. Here we use LSTM to parse a review and identify the associated sentiment (positive/negative)

A fun application of LSTMs to generate metal lyrics from all the lyrics ever written by that band. We explore both character based and word based sequence generation techniques. Play with this notebook to generate lyrics for your metal band. Mostly it will be gibberish, but I am sure there will be some interesting phrases