Back To Schedule
Wednesday, October 16 • 2:00pm - 2:40pm
Reinforcement Learning for crop-management: Sequential decision-making under uncertainty FULL

Log in to save this to your schedule, view media, leave feedback and see who's attending!

Feedback form is now closed.
Limited Capacity full

Reinforcement Learning (RL) is a branch of Machine Learning that is different from classical Supervised or Unsupervised Learning. Reinforcement Learning problem is for an agent to find an optimal policy, i.e. what action should be chosen for any given system's state, and to maximize a goal along a series of actions in a partially known of fully unknown stochastic environment. RL will be presented as a natural tool for agricultural decision-making. RL agricultural use cases come mainly from the 90’s where available methods and data context were different. The utilization of both new data sources and RL methods should be re-investigated for agricultural decision-making, being relevant to climate-resilience and sustainable-intensification global necessities calling for dynamic policies. In that perspective, RL problems and solving methods will be introduced and illustrated through an agricultural example. Promising state-of-the-art methods combined with a new data context, challenges ahead will be discussed to raise Data-Driven Agronomy Community's interest. The session will examine, in part, how RL approaches can leverage existing modeling techniques and agronomic data commonly collected by CGIAR (such as through the AgroFIMS tool developed by the Platform for Big Data) as inputs in a dynamic decision engine.

avatar for Philippe Preux

Philippe Preux

Professor, Université de Lille & Inria
“Philippe Preux, professor at Université de Lille, France, co-funder and head of INRIA SequeL [Sequential Learning] team, one of the leading Reinforcement Learning laboratories. Since past 12 years SequeL has been contributing to RL literature with major publications and actively... Read More →
avatar for Romain Gautron

Romain Gautron

Data Scientist, CIAT

Wednesday October 16, 2019 2:00pm - 2:40pm GMT+05
Breakout 4 (New Sahel) ICRISAT-HQ