In this video, Siraj Raval explores how key reinforcement learning algorithms help explain how the human brain works, specifically through the lens of the neurotransmitter known as ‘dopamine’.

These algorithms have been used to help train everything from autopilot systems for airplanes, to video game bots. TD-Learning, Rescorla-Wagner, Kalman Filters, and Bayesian Learning, all in one go!

By leveraging powerful prior knowledge about how the world works, humans can quickly figure out efficient strategies in new and unseen environments.

Currently, even state-of-the-art Reinforcement Learning algorithms typically don’t have strong priors and this is one of the fundamental challenges in current research on Transfer Learning.

Related Links

Check out this great demo of the new way to configure message routing in Azure IoT Hub using Device Twin properties and the updated UI in the Azure portal.

Paul Montgomery, Senior Software Engineer in the Azure IoT Team shows us how this all works using an ESP32 based microcontroller, a temperature sensor, a glass of hot water and another glass of iced water.

While at NDC in Sydney, Carl and Richard talked to Joe Albahari about using LINQPad to create neural nets from scratch.


LINQPad is an interactive development environment for .NET – originally focused on helping you build LINQ expressions. But as Joe explains, it can be used for all sorts of interactive coding experiences – including learning to build neural networks. Joe talks through the fundamentals of neural nets and what it’s like to build neural nets yourself. Even if you move on to more advanced machine learning tooling, learning the fundamentals are useful!

Joe Albahari is an O’Reilly author and the inventor of LINQPad. He’s written seven books on C# and LINQ, including the upcoming “C# 7.0 in a Nutshell”. He speaks regularly at conferences and user groups, and has been a C# MVP for nine years running.

Press the play button below to listen here or visit the show page.

Siraj Raval explores the world of automated training with reinforcement learning with a few lines of Python code!

In this video, he demonstrates how a popular reinforcement learning technique called “Q learning” allows an agent to approximate prices for stocks in a portfolio. The literature of reinforcement learning is incredibly rich. There are so many concepts, like TD-Learning and Actor-Critic for example, that have real-world potential.

On the podcast Andy Leonard and I create, we love experimenting around here and examining the resulting data. After all, we are Data Driven: not just in name but also in spirit.In this webinar Andy recorded, he also streamed it live on our Facebook page.

We thought it was good enough to share with our larger audience here.Let us know what you think. Both Frank and Andy have been recording/streaming their live events and we’re curious to hear what you have to say about this innovation in how we podcast.

Press the play button below to listen here or visit the show page at DataDriven.tv