Building DeepNets using Keras
Submitted by Anuj Gupta (@anuj-gupta) on Tuesday, 24 May 2016
Section: Workshop Technical level: Intermediate
In todays world many ML teams have started to look towards DeepNets as potential models to build intelligent system. While DeepNets hold lot of promise, building them from scratch can be very time consuming. What is needed is elaborate library that will facilitate quick experimentation. Thats where Keras comes handy. Its moto is - “Being able to go from idea to result with the least possible delay is key to doing good research.”
- It is a a python library that provides a clean and convenient way to create a range of deep learning models on top of Theano or TensorFlow.
- Allows for easy and fast prototyping (through total modularity, minimalism, and extensibility).
- Supports both convolutional networks and recurrent networks, as well as combinations of the two.
- Runs seamlessly on CPU and GPU.
- Has a strong community - http://kerasteam.slack.com/
Keras was started in march, 2015 by François Chollet, a google engineer. Just last month its version 1.0 was released.
In this session we will take a deep dive in Keras, get our hands dirty by building some DeepNets using Keras API.
- Introduction to Keras
- Using Keras build/code popular neural networks - Feedforward, Convolutional, Recurrent - LSTM, GRU
- See how to Word Embedding to initialise Enbedding Layer
- Use Theano / TensorFlow as backend
The aim of the workshop is to introduce Keras (http://keras.io/) for building Deep Nets.
I currently work as a pricipal ML researcher at Airwoot(Now acquired by Freshdesk), building inteligent applications using NLP + Deep Learning.
Prior to joining industry I was a part of Data-Science group at IIT Delhi and research scholar with theory group at IIIT - hyderabad.
You can find more about me on