Date |
Topic |
Readings and assignments
|
| Wed 8/22 | Course overview and introduction. | Short overview of advances
in AI Recent progress in deep learning |
| Mon 8/27 | First machine learning algorithm: gradient descent |
Introduction
to gradient descent Gradient descent for linear regression |
| Wed 8/29 | Logistic regression and classification learning |
Gradient
descent for logistic regression |
| Wed 9/5 | Introduction to neural networks and the perceptron learning rule | History
of deep learning Perceptron implementation |
| Mon 9/10 | Backpropagation algorithm for training feedforward multi-layer neural networks | Backpropagation
algorithm Backpropagation implementation |
| Wed 9/12 | Training, overfitting, cross-validation |
|
| Mon 9/17 | Overview of deep learning |
|
| Wed 9/19 | Search in AI: backtracking and branch-and-bound search |
|
| Mon 9/24 | Constraint propagation (forward checking) as an approach
to improving backtracking search Problem relaxation as an approach to computing bounds for branch-and-bound search |
|
| Wed 9/26 | State-space search in AI: problem representation, search strategies and heuristics. | |
| Mon 10/1 | A* search algorithm: admissible and consistent heuristics. | |
| Wed 10/3 | A* search algorithm and extensions | Python implementation of A*
for sliding-tile puzzles Python implementation of A* for pathfinding in mazes |
| Mon 10/8 | Game-tree search: minimax and alpha-beta pruning |
Minimax
search with alpha-beta pruning |
| Wed 10/10 | Game-tree search continued: static evaluations functions,
iterative deepening, node ordering |
More
about game-tree search |
| Mon 10/15 | Game-tree search continued:
state-of-the-art approaches, AlphaGo, AlphaGoZero, and
AlphaZero |
Deep learning and Monte-Carlo search
for AlphaGo Details of AlphaGoZero |
| Wed 10/17 | Planning and search in stochastic domains: Markov decision
processes |
|
| Mon 10/22 | Planning and search in stochastic domains: Value iteration
and policy iteration |
|
| Wed 10/24 | Planning as state-space search. Discussion of programming
assignment. |
|
| Mon 10/29 | Introduction to reinforcement learning: Q-learning |
Value
iteration, Q-learning, Python implementation for balancing
cartpole Reinforcement learning algorithms Deep reinforcement learning |
| Wed 10/31 | Review |
|
| Mon 11/5 | EXAM |
|
| Wed 11/7 | Game theory and multi-agent AI: Nash equilibrium,
iterative elimination of dominated strategies |
|
| Mon 11/12 | Review of probability theory and Bayes rule |
|
| Wed 11/14 | Naive Bayes algorithm |
Different
types of machine learning |
| Mon 11/19 | Review |
|
| Mon 11/26 | Nearest neighbors algorithm for classification and
regression Unsupervised learning and k-means clustering |
K-Nearest
neighbors Unsupervised learning and k-means clustering |
| Wed 11/28 | Decision tree classifiers and boosting |
Decision
trees |
| Mon 12/3 |
||
| Wed 12/5 |
Topics
for further research |
|
| Tues 12/11 | Final
exam 3 - 6 |