Pieter Abbeel
Professor of Electrical Engineering and Computer Sciences · UC Berkeley · ACM Prize Laureate
Pieter Abbeel is Professor of EECS at UC Berkeley, director of the Berkeley Robot Learning Lab and an ACM Prize in Computing recipient for deep reinforcement learning.
- lectures
- 25
- series
- 1
- hours
- 35
Start here
Series and courses
UC Berkeley CS 188: Introduction to Artificial Intelligence
Open the course
- 1Introduction to Artificial Intelligence
- 2Uninformed Search
- 3A* Search and Heuristics
- 4Constraint Satisfaction Problems (CSPs) Part 1/2
Show 21 moreShow fewer
- 5Constraint Satisfaction Problems (CSPs) Part 2/2
- 6Search with Other Agents: Minimax
- 7Search with Other Agents: Expectimax, Utilities
- 8Markov Decision Processes (MDPs) Part 1/2
- 9Markov Decision Processes (MDPs) Part 2/2
- 10Reinforcement Learning Part 1/2
- 11Reinforcement Learning Part 2/2
- 12Probability
- 13Bayes' Nets: Representation
- 14Bayes' Nets: Independence (D-Separation)
- 15Bayes' Nets: Inference (Variable Elimination)
- 16Bayes' Nets: Sampling
- 17Decision Networks and Value of Perfect Information (VPI)
- 18Hidden Markov Models
- 19Particle Filtering
- 20Machine Learning: Naive Bayes
- 21Machine Learning: Perceptrons and Logistic Regression
- 22Machine Learning: Optimization and Neural Networks
- 23Machine Learning: Neural Networks and Decision Trees
- 24Robotics, Language, and Vision
- 25Advanced Topics and Course Summary
In these lectures
- Markov Decision Processes (MDPs) generalize search problems to non-deterministic environments with probabilistic transition functions and reward structures.Markov Decision Processes (MDPs) Part 1/2
- Planning agents use a world model and goal tests to evaluate action sequences, unlike reflex agents which react purely to current percepts.Uninformed Search
- Reinforcement learning optimizes agent behavior by interacting with an environment and learning from observed rewards and next states.Reinforcement Learning Part 1/2
Draws on
People cited in the lectures, at the moment of the citation.
- Dan KleinInstructor of CS 188 at UC Berkeley
- Andrew NgProfessor at Stanford, cited for his lab's work on snake robots.
- Lee SedolProfessional Go player defeated by AlphaGo.
- Andrey Markov1856-1922, namesake of Markov processes
- Peter StoneProfessor at the University of Texas, Austin, cited for his work on RoboCup.
- Russ TedrakePhD thesis at MIT, cited for the toddler robot example.
About
Pieter Abbeel is Professor of EECS at UC Berkeley, director of the Berkeley Robot Learning Lab and an ACM Prize in Computing recipient for deep reinforcement learning. His CS 188 lectures teach undergraduate AI, from A* search and d-separation in Bayes' nets to value iteration and Q-learning.
Taught with
Same subject, AI & Machine Learning
Connections
30Their places, courses, books and lectures.
Concept1
Lectures25
- Markov Decision Processes (MDPs) Part 1/2
- Uninformed Search
- Reinforcement Learning Part 1/2
- A* Search and Heuristics
- Search with Other Agents: Minimax
- Bayes' Nets: Independence (D-Separation)
Show 19 moreShow fewer
- Bayes' Nets: Sampling
- Bayes' Nets: Representation
- Probability
- Hidden Markov Models
- Machine Learning: Optimization and Neural Networks
- Constraint Satisfaction Problems (CSPs) Part 1/2
- Decision Networks and Value of Perfect Information (VPI)
- Markov Decision Processes (MDPs) Part 2/2
- Introduction to Artificial Intelligence
- Bayes' Nets: Inference (Variable Elimination)
- Search with Other Agents: Expectimax, Utilities
- Constraint Satisfaction Problems (CSPs) Part 2/2
- Machine Learning: Perceptrons and Logistic Regression
- Machine Learning: Naive Bayes
- Particle Filtering
- Reinforcement Learning Part 2/2
- Robotics, Language, and Vision
- Machine Learning: Neural Networks and Decision Trees
- Advanced Topics and Course Summary


