Webinar: Beyond forecasting and predictions - Using Machine Learning to Control Real-World Systems

Webinar Recording

About the Webinar: 

Until a few years ago, introductory courses in machine learning only covered supervised and unsupervised algorithms. Most material would simply mention that a third type known as reinforcement learning (RL) existed. In recent times, RL has gained greater prominence by defeating human experts in many games, including Atari, Go, and DoTA.

Nevertheless, there is still a lack of clarity in popular understanding of reinforcement learning, where it applies, how it works, and how it applies to tasks other than playing games. This talk will attempt to deconstruct the concepts of RL and present them in intuitive form. It will also cover several case studies from the application of RL to real-world problems.

Key Takeaways from the Webinar:

• Where does RL fit in the context of other machine learning methods?
• How does RL work?
• Where can/should RL applied?
• Examples from real-world problems in transportation, logistics, supply chain.

Who is this Webinar for?

ML practitioners, business domain experts, and technical managers who wish to obtain conceptual clarity about reinforcement learning as it applies to their fields of interest. In this talk, I will not be going deep into mathematics or code.

Speaker: 

Dr. Harshad Khadilkar, Scientist(Research Division), TCS

Harshad Khadilkar is a scientist with TCS Research as well as a visiting associate professor at IIT Bombay. He holds a BTech from IIT Bombay and an SM and PhD from the Massachusetts Institute of Technology. His research interests include control and optimisation problems in distributed networks and industrial systems.