Presented By O’Reilly and Intel Nervana
Put AI to work
September 17-18, 2017: Training
September 18-20, 2017: Tutorials & Conference
San Francisco, CA

Introduction to reinforcement learning

Marcos Campos (Bonsai)
9:00am–12:30pm Monday, September 18, 2017
Implementing AI
Location: Yosemite BC
Secondary topics:  Algorithms, Data science and AI, Transportation and autonomous vehicles
Average rating: **...
(2.33, 9 ratings)

Prerequisite Knowledge

  • General experience with programming
  • Familiarity with supervised, unsupervised, and reinforcement learning

Materials or downloads needed in advance

  • A laptop

What you'll learn

  • Understand the basics of reinforcement learning
  • Learn how to build and train an AI model using reinforcement learning techniques


Marcos Campos offers an overview of reinforcement learning, walking you through the various classes of reinforcement learning algorithms, the types of problems that can be solved with this technique, and how to build and train AI models using reinforcement learning and reward functions.


Types of learning

  • Supervised learning
  • Unsupervised learning
  • Reinforcement learning

Conceptual overview of reinforcement learning

  • Use cases best suited for reinforcement learning
  • Challenges of reinforcement learning (dynamic programming problem, reward functions, and scalability)
  • Strategies to make reinforcement learning problems easier to solve (shaping, curriculum learning, apprenticeship learning, building blocks, and multiconcepts)

Classes of reinforcement learning algorithms

  • On-policy versus off-policy methods
  • Model-based algorithms
  • Actor-Critic
  • Q-learning
  • Parallel algorithms

A step-by-step demonstration of building and training an AI model using reinforcement learning

  • Use the Bonsai Platform to leverage reinforcement learning techniques and train AI models
Photo of Marcos Campos

Marcos Campos


Marcos Campos is head of artificial intelligence at Bonsai, where he leads the artificial intelligence team and is helping create a next-generation AI platform that will unlock intelligence across different types of products and applications. Marcos has been working at making machine learning and artificial intelligence accessible to a large community of users for over a decade. Previously, he set the vision and strategy roadmap for machine learning at Oracle and Uber, focusing on creating scalable and easy-to-use platforms. He has been awarded over 23 patents for this work.

Comments on this page are now closed.


09/05/2017 11:21am PDT

Could I get Bonsai Platform before conference?
I concerned with any troble about install.