Keras in Motion
Dan Van Boxel
55 exercises
  • June 2017
  • 2h 5m

A great introduction to using Keras for deep learning.

Daniel Williams, Software Professional

See it. Do it. Learn it! Keras in Motion introduces you to the amazing Keras deep learning library through high-quality video-based lessons and built-in exercises, so you can put what you learn into practice.

Keras in Motion teaches you to build neural-network models for real-world data problems using Python and Keras. In nearly 3 hours of hands-on, practical video lessons, you'll apply Keras to common machine learning scenarios, ranging from regression and classification to implementing Autoencoders and applying transfer learning. In each crystal-clear video module, you'll put your new knowledge into practice, as you teach your network to recognize text and even create an algorithm for a self-driving car!

Table of Contents detailed table of contents

Installation and Basics

What is Keras


Fitting Lines

Linear Regression in Keras

Predicting Categories

Logistic Regression in Keras

Font Recognition

Exploring Font Data

Neural Networks Review

Neural Networks in Keras

Finding Simple Features

Convolutional Neural Networks in Keras

Visualization and Wrap Up

Self Driving Car

Exploring Self Driving Car Data

Simple Steering Model

Convolutional Steering Model

Combined Steering Throttle Model

Summary and Extensions


Introduction to Autoencoders

Autoencoding MNIST Digits

Transfer Learning

Course Review

Converting Keras

Converting Keras1 to Keras2

About the subject

Keras is a Python library designed to take the stress out of deep learning. The Keras library provides a library of high-level building blocks on top of the low-level features of the TensorFlow and Theano machine learning frameworks. In Keras, you define deep learning models without specifying the detailed mathematics and other mechanics, so you can focus on what you want to accomplish. Built with experimentation and prototyping in mind, Keras has a super friendly API and an intuitive Python-based coding style. With over 50,000 users, Keras is the perfect choice for any developer working with data.


Written for intermediate-level data scientists, developers, and machine learning engineers. Code examples are in Python.

What you will learn

  • Regression and classification problems
  • Using neural networks for image processing
  • Building autoencoders
  • Designing and implementing a self-driving car
  • Hands-on coding with practical exercises and example

About the instructor

Dan Van Boxel is an engineer and data scientist with a background in both engineering and mathematics. On his livestream, Dan demonstrates a different machine learning library, method, or model weekly.

liveVideo $49.99

Makes deep learning much more straight forward.

Peter Hampton, AI Researcher, Ulster University

The instructor is capable of breaking complex concepts into easily understandable examples.

Gustavo Patino, Assistant Professor, Oakland University William Beaumont School of Medicine