Applying Reinforcement Learning in Industry

💥This is our fifth meetup for 2021💥

In this event Pavlos Mitsoulis returns to talk to us about applying reinforcement learning techniques in the industry

Grab your beer 🍺 at home and join us virtually through the zoom link. We look forward to seeing you there.

Abstract

Most people are familiar with or have heard that Reinforcement Learning (RL) has been employed successfully on games. The biggest example is AlphaGo, developed by DeepMind, the first computer program to defeat the strongest Go player in history. The process of learning continuously from the feedback of the environment sounds very appealing for solving e-commerce use cases. Why we don't see many applied RL use cases? In this talk, Pavlos will explain how Contextual Multi-Armed-Bandits, which is a family of RL algorithms, can be exploited for applied use cases. Contextual MABs are able to optimize directly business KPIs (Conversion, CTR, etc), no training data is required and are use case agnostic.

About the speaker

Pavlos Mitsoulis is a Staff Data Scientist leading the Reinforcement Learning team at Expedia Group and the co-creator of Sagify, an open source MLOps CLI tool to easily train, tune and deploy ML models on Sagemaker by implementing just a train and a predict function. His interests are around MLOps and Contextual Multi-Armed-Bandits these days.

Code of conduct

https://pydata.org/code-of-conduct/