Deep Learning with (Keras + Tensorflow)
Deep Learning is a flavour of Machine Learning which make use of Deep Neural Networks for better and more sophisticated machine learning models.
However, the art of building Deep Learning models is still perceived as rocket science by most people.
This talk would explain the audience about how to spin up deep learning models very easily with Keras, using Tensorflow a the backend.
Target Audience: Data Scientists, or people who want to get started in the Deep Learning domain
Prerequisites: Knowledge of Python. Basic understanding of Map-Reduce(but this is not a strict pre-req)
- Introduction to a basic sequential neural net model
- A quick coding up of the model
- Understanding and appreciaitng the simplicity of Keras, and how Tensorflow acts as it’s backend
- A brief understanding of how Tensorflow works, under the hood
- Improving up on the model, thus spinning up a deep neuralnet, along with understanding(and appreciating) how each layer works and how the final prediction is done.
I work as a Senior Software Engineer - Data Science at Zomato. I work on several ML and data science problems at work and as side projects, specializing in recommender systems and neural nets.