Teacher(s)
Language
French
Prerequisites
LEPL1402: Programming in a high-level language
Main themes
- Problem solving through research: problem formulation, informed and uninformed search strategies, local search, performance evaluation and cost estimation, practical applications.
- Constraint satisfaction: problem formulation, monitoring and propagation of constraints, use cases and various applications.
- Games and adversarial search: Minimax algorithm, Alpha-Beta pruning, Monte-Carlo Tree Search, and application examples.
- Logic: propositional and first-order logic, knowledge representation, inference, reasoning, and applications in various contexts.
- Decision making: simple and complex decisions, collective decisions, probability theory, decision networks, Markov decision-making processes, game theory, multi-agent systems.
- Learning: introduction to supervised learning, decision trees, reinforcement learning, regression, and practical applications.
Learning outcomes
At the end of this learning unit, the student is able to : | |
With respect to the learning outcomes framework (AA) of the Bachelor's program in Computer Science, this course contributes to the development, acquisition, and assessment of the following learning outcomes:
With respect to the learning outcomes framework (AA) of the Bachelor's program in Engineering Sciences, Civil Engineering orientation, this course contributes to the development, acquisition, and assessment of the following learning outcomes:
Students who successfully complete this course will be able to:
Students will also acquire methodological and operational skills. In particular, they will develop the ability to:
|
|
Content
The course is organized around four main themes:
- Search strategies and problem solving.
- Games, decision-making, and multi-agent systems.
- Logic, knowledge representation, and automated reasoning.
- Machine learning, from its fundamental principles to deep learning.
- General introduction to Artificial Intelligence (intelligent agents, rationality, agent-environment interaction, task environments, and the main paradigms of AI).
- Uninformed search strategies (problem formulation, state spaces, breadth-first search, depth-first search, uniform-cost search, depth-limited search, and iterative deepening).
- Informed search strategies (heuristic functions, greedy best-first search, A* search, admissibility, consistency, optimality, and heuristic design).
- Adversarial search and Monte Carlo Tree Search (game trees, Minimax, Alpha-Beta pruning, heuristic evaluation functions, MCTS, and Upper Confidence Bounds for Trees).
- Simple decision-making under uncertainty (probabilistic reasoning, preferences, utility functions, expected utility, decision networks, and value of information).
- Complex and sequential decision-making (Markov Decision Processes, policies, value functions, Bellman equations, policy iteration, and value iteration).
- Multi-agent systems (cooperative and competitive interactions, normal-form games, strategies, utilities, dominant strategies, Nash equilibria, coordination, and collective decision-making).
- Propositional logic and reasoning (knowledge-based agents, syntax and semantics, logical equivalence, validity, satisfiability, entailment, conjunctive normal form, model checking, and resolution).
- First-order logic and automated reasoning (predicates, functions, variables, quantifiers, unification, substitution, forward and backward chaining, and resolution).
- Supervised machine learning (classification, regression, linear and logistic regression, loss functions, gradient descent, backpropagation).
- Deep learning (multilayer neural networks, activation and loss functions, weight initialization, universal approximation theorem).
Teaching methods
The course combines lectures, group work, and individual work.
Evaluation methods
- For the first session, the assessment consists of three parts:
- Individual comprehensive final exam (50%). The use of AI is not permitted for this assessment. The exam may be answered in English or French.
- Individual exercises (20%): exercises to be completed individually will be assigned throughout the session. The use of AI is not permitted for this assessment.
- Team project (30%): implementation of an agent based on reinforcement learning. The use of AI is permitted for this assessment, in accordance with the rules defined by the faculty.
- Important: to obtain a passing grade for the course (>= 10/20), students must obtain a minimum score of 20/50 on the final exam. If this condition is not met, only the exam grade will be taken into account. The final exam assesses critical skills that are not covered by the other assessments and are essential for successfully completing the course unit.
For the second session, the exam and individual exercises may be retaken under the same weighting and conditions. The project grade is retained across all sessions and the project cannot be retaken.
In the event of re-enrollment, all assessments must be taken again.
Online resources
Bibliography
Part of the course content is based on the following book:
- Artificial Intelligence: A Modern Approach (4th Edition): https://aima.cs.berkeley.edu/
Faculty or entity
Programmes / formations proposant cette unité d'enseignement (UE)
Title of the programme
Sigle
Credits
Prerequisites
Learning outcomes
Master [120] in Chemical and Materials Engineering
Master [120] in Civil Engineering
Master [120] in Biomedical Engineering
Specialization track in Computer Science
Master [120] in Mechanical Engineering
Master [120] in Electrical Engineering
Master [120] in Physical Engineering
Bachelor in Computer Science
Master [120] in Electro-mechanical Engineering
Master [120] in Data Science Engineering
Master [120] in Data Science: Information Technology
Master [120] in Energy Engineering
Mineure Polytechnique