Simons Institute Courses

Language as a Scaffold for Reinforcement Learning

Explore how natural language can enhance reinforcement learning, covering topics like instructions as observations, task decomposition, and learning from demonstrations.

• YouTube
• 31 minutes
• Self-Paced
• Free Video

Learning Exploration Strategies with Meta-Reinforcement Learning

Discover meta-reinforcement learning strategies for efficient exploration in complex environments, focusing on posterior sampling and task-relevant information.

• YouTube
• 31 minutes
• Self-Paced
• Free Video

Discussion - Offline Reinforcement Learning

Explore offline reinforcement learning, covering data sets, optimism under uncertainty, state coverage, and policy evaluation. Gain insights into pragmatic and conceptual approaches.

• YouTube
• 40 minutes
• Self-Paced
• Free Video

Offline Deep Reinforcement Learning Algorithms

Explore offline deep reinforcement learning algorithms, their challenges, and solutions like conservative Q-learning, with insights on modern machine learning and overcoming distributional shift.

• YouTube
• 32 minutes
• Self-Paced
• Free Video

MOPO - Model-Based Offline Policy Optimization

Learn about MOPO, a model-based approach for offline reinforcement learning, addressing challenges in distributional domain shift and improving policy optimization.

• YouTube
• 38 minutes
• Self-Paced
• Free Video

Berkeley in the 80s - Andrew Yao's Turing Award-Winning Research

Explore Andrew Yao's groundbreaking research at Berkeley, his journey in theoretical computer science, and insights for aspiring researchers in this engaging interview.

• YouTube
• 42 minutes
• Self-Paced
• Free Video

Insights on Gradient-Based Algorithms in High-Dimensional Learning

Explore gradient-based algorithms in high-dimensional learning, covering statistical physics solutions for performance analysis in nonconvex settings like spiked mixed matrix-tensor model, perceptron, and phase retrieval.

• YouTube
• 1 hour
• Self-Paced
• Free Video

A Few Challenge Problems from Robotics

Explore challenging robotics problems, including backflips, models, perception, and optimization. Gain insights into reinforcement learning, state estimation, and physics-based approaches for advanced robotic systems.

• YouTube
• 1 hour 11 minutes
• Self-Paced
• Free Video

Simulation Methodology - An Overview

Explore simulation methodology, covering numerical integration, Monte Carlo methods, and output analysis. Learn key concepts and techniques for tackling high-dimensional problems in computational science.

• YouTube
• 1 hour 7 minutes
• Self-Paced
• Free Video

Planning and Markov Decision Processes - Part 1

Explore fundamental concepts of planning and Markov Decision Processes, including control objectives, randomizing policies, and basic methods for reinforcement learning theory.

• YouTube
• 1 hour 17 minutes
• Self-Paced
• Free Video

Stochastic Programming Approach to Optimization Under Uncertainty - Part 2

Explore advanced optimization techniques for uncertain scenarios, covering dynamic programming, discretization, cutting planes, and multistage problems in reinforcement learning theory.

• YouTube
• 1 hour 9 minutes
• Self-Paced
• Free Video

Statistical Considerations in Reinforcement Learning - Part 1a

Explore statistical considerations in reinforcement learning, covering inverse RL, high-stakes problems, and asymptotic frameworks for parameter estimation and uncertainty quantification.

• YouTube
• 1 hour 14 minutes
• Self-Paced
• Free Video

The AI Economist- Improving Equality and Productivity with AI-Driven Tax Policies

Explore AI-driven tax policies for improved economic equality and productivity, featuring deep learning, reinforcement learning, and multi-agent simulations in a simplified economic model.

• YouTube
• 1 hour 14 minutes
• Self-Paced
• Free Video

Batch Offline Reinforcement Learning - Part 1

Explore offline reinforcement learning theory, covering key concepts, challenges, and evaluation methods for developing AI systems using pre-collected data.

• YouTube
• 59 minutes
• Self-Paced
• Free Video

Online Learning in Markov Decision Processes - Part 2

Explore advanced concepts in Markov Decision Processes, including adversarial scenarios, regret analysis, and algorithms for online learning in dynamic environments.

• YouTube
• 1 hour 13 minutes
• Self-Paced
• Free Video