Anthill Inside 2017

On theory and concepts in Machine Learning, Deep Learning and Artificial Intelligence. Formerly Deep Learning Conf.

Keras: Deep Learning for Python

Submitted by Fariz Rahman (@farizrahman4u) on Friday, 16 June 2017

videocam_off

Technical level

Advanced

Section

Full talk

Status

Submitted

Vote on this proposal

Login to vote

Total votes:  +9

Abstract

An introduction to the python deep learning library Keras, the philosophy behind Keras, building and training trivial as well as complex models such as GANs, RL, etc using Keras, deploying a Keras model in a production environment, and the future of Keras. Intended audience: Basic algebraic skills and python experience.

Outline

  • Introduction to DL and DL libraries - Theano, Torch, Tensorflow, Keras
  • Keras and the democratization of DL - the philosophy of Keras, importance of API simplicity, Keras as a spec rather than a project
  • Building and training basic Keras models - working with datasets, optimizing input pipeline, using pre-trained models
  • Advanced models : GANs, Seq2seq, RL.
  • Deployment : Deploying a keras model on servers and mobile apps.
  • Future of Keras: Roadmap, community, …

Requirements

Laptop

Speaker bio

  • Top Keras contributor. (contributions)
  • Keras community member. Respnosibilities include maintaining and contributing to the core Keras repo and Keras community contributions repo, responding to user queries and issues, helping with roadmaps, and suggest new directions for the project.
  • Have written various popular libraries that are part of the Keras ecosystem: qlearning4k for RL, RecurrentShop for RNNs, Seq2seq and a couple of others
  • Working as Machine Learning Engineer at (datalog.ai)(www.datalog.ai)
  • Currently authoring a book on deep learning using Keras with Packt Publishing

Comments

  • 1
    Sandhya Ramesh (@sandhyaramesh) Reviewer a year ago

    Hi Fariz, in order to evaluate your proposal, we need your slide deck and a two minute self recorded preview video explaining your talk. Please edit this proposal and submit them. Thanks.

Login with Twitter or Google to leave a comment