Stochastic Optimization and Reinforcement Learning

An introduction about how to train computers to interact wisely with a stochastic process scenario to maximize certain benefit/reward

Duration

8 to 12 h

Syllabus

Markov decision processes, dynamic programming and related algorithms

Fundamental concepts in Reinforcement Learning:

• temporal difference methods,
• approximate dynamic programming
• deep reinforcement learning

Prerequisites

“Foundations of Data Science”, highly advisable “Deep Learning and applications”

Credential

Part of ‘Stochastic models and optimization’ in MSc in Data Science, also taught in Data Science Center workshops