مخطط الموضوع
-
“An online test is scheduled for Friday, December 19 at 20h:00 .”
-
The professor in charge of the scale: Dr.MENAD
E-mail: menad.h92@gmail.com
LinkedIn: Hanane MENAD
Module title: Basics of Artificial Intelligence
Unit title: UEM1
Coefficient: 3
Credit: 5
Hourly volume: STV (60h); C(1h30); TD(1h30); TP(1h30)
Assessment method: Exam (60%), continuous assessment (40%).
-
-
-
-
مفتوح: الجمعة، 19 ديسمبر 2025، 8:00 PMمغلق: الجمعة، 19 ديسمبر 2025، 8:05 PM
-
-
-
Course objective:
The course objective is the acquisition of concepts and basic techniques of artificial intelligence fields. To be able to model and solve real problems using the appropriate techniques of artificial intelligence.
Recommended prior knowledge:
Algorithms and data structures, Graph theory, Language theory, Probability theory, Propositional logic.
-
-
This chapter introduces the fundamentals of Artificial Intelligence, including its definition, key concepts (Machine Learning, Neural Networks, NLP, Computer Vision, Expert Systems, and Optimization), and main applications in healthcare, finance, education, transport, and more. It also covers the history and milestones of AI, from Turing’s theoretical machine to modern AI systems like GPT, highlighting the evolution of algorithms, training methods, and ethical considerations in AI development
-
This chapter introduces the concept of intelligent agents, their properties (autonomy, proactivity, flexibility, social ability, situatedness), and types (autonomous, reactive, hybrid). It also explores environment characteristics, such as accessibility, determinism, discreteness, and dynamics, and highlights how agents perceive, decide, and act in various settings. Practical examples include thermostats, printers, robots, and smart cars.
-
This chapter focuses on problem representation in AI, highlighting how complex problems can be formalized, modeled, and decomposed into simpler sub-problems. It covers state graphs, AND/OR graphs, and problem formalization using inputs, outputs, rules, and strategies. The chapter emphasizes the importance of clear representation for understanding, solving, and communicating problems effectively, and discusses algorithmic complexity, decision-making, and applications in planning, optimization, and expert systems.
-
This chapter presents uninformed (blind) search algorithms used to explore state spaces without domain-specific knowledge. It covers BFS (Breadth-First Search), DFS (Depth-First Search), IDS (Iterative Deepening Search), Two-Way Search, and UCS (Uniform Cost Search). Students will learn the principles, operation, and examples of each algorithm, understand their advantages and limitations, and see how they are applied to find paths or solutions in graphs and trees.
-
This chapter introduces informed (heuristic) search techniques, which use evaluation functions to guide the exploration of the search space efficiently. It covers Best-First Search, Greedy Search, and advanced algorithms such as A*, SMA*, IDA*, Weighted A*, and Theta*. Students will learn how these methods optimize pathfinding and decision-making, their advantages, limitations, and practical applications in games, planning, and large-scale problem-solving.
-
This chapter presents Constraint Satisfaction Problems (CSP) and local search algorithms, two core approaches in AI for solving complex optimization problems.
Constraint search (CSP) and local algorithms are used in AI to solve complex problems. CSP models problems with variables, domains, and constraints, while local algorithms like Simulated Annealing and Hill Climbing iteratively improve solutions to find optimal or near-optimal results.
-
Genetic Algorithms (GA) are optimization methods inspired by natural selection. They evolve a population of solutions through selection, crossover, and mutation to find the best solution. GAs are used in optimization, scheduling, routing, and AI. They explore solutions widely but can be computationally expensive.
-
In this chapter, game-solving algorithms in AI include Minimax for two-player games, Expectimax for games with chance, and Monte-Carlo Tree Search (MCTS). Minimax maximizes a player’s gain while minimizing the opponent’s, with alpha-beta pruning to cut unnecessary branches. Expectimax adds probabilistic chance nodes, while MCTS uses random simulations to explore large or complex game spaces efficiently.
-
In this chapter, Markov chains are introduced as models for stochastic processes where the future depends only on the current state. They are applied in Hidden Markov Models (HMMs) for language, speech, and activity recognition, and in Markov Decision Processes (MDPs) for reinforcement learning, guiding sequential decision-making under uncertainty with states, actions, probabilities, rewards, and discount factors.
-
In this chapter, neural networks are introduced as brain-inspired models that learn patterns from data, used in image recognition, NLP, and AI tasks. Fuzzy logic handles uncertainty by allowing partial truth values, applied in automation, AI, medicine, and robotics.
-
-
Exam ملف PDF