Intelligence artificielle

This course will allow students to learn about the basics of artificial intelligence. Four themes will be covered,

• Search and planning; covering topics such as informed and uninformed search, local research, games and adversarial search.
• Probabilistic reasoning; covers topics such as Bayesian networks and Markov models.
• Decision-making under uncertainty; with topics like Markov Decision Processes and Reinforcement Learning.
• machine learning; with topics like naive Bayes, regression, perceptrons and neural networks.

With this course, students should have enough technical knowledge and skills to work on AI-related projects and successfully complete AI-related courses in the Master CS program at ULB and other universities.

Required and Corequired knowledge and skills

Programming, algorithmics and standard mathematics knowledge obtained in the first Bachelor year.

Required and corequired courses

Theoretical sessions (24h) and exercises (24h) and five (one optional) projects (60h).  

• The theory session is each time 1 hour followed by a 1-hour exercises session, and this twice per week.  
• The exercises are organised after each 1hour theory session, where students will solve Ai problems related to each part of the course.   
• The project consists of five programming assignments that will be provided during the year at different intervals.  They will cover the main themes of this course. 
   

Contribution to the teaching profile

 

References, bibliography, and recommended reading

This course is directly based on AI - a Modern Approach, 4th edition. There are both an English and French version of this book.  You can also get access to an online copy via this link.

the ULB library also has 4-5  copies of this book available.

 

Course notes

• Université virtuelle

Additional information

 

Contacts

 

Campus

Plaine

Method(s) of evaluation

• written examination
• Project

written examination

Project

• The exam consists of a series of exercises/problems/questions seen in the practical sessions of the course. On UV, a set of exercises with solutions is provided which contains sample exam questions.
• Projects are implementations of different parts of the course. Last year, this consisted of implementations in a PacMan environment of search algorithms, adversary algorithms, probabilistic models, reinforcement learning agents, and machine learning agents. While the type of projects will be the same, the AI environment may change.

Mark calculation method (including weighting of intermediary marks)

1. The project grade will be the average of the 4 best results obtained for the 5 projects
2. The final grade of the course is calculated as follows.

• If the exam score or the total score of the projects is lower than 9/20, then the lowest grade is used for the final grade of this course.
• Otherwise the score is a weighting between the exam results and the projects, calculated as follows
  • 60% of the final grade is obtained on the exam and
  • 40% of the total grade obtained on the projects

Language(s) of evaluation

• french
• (if applicable english, Dutch )