Introduction to Artificial Intelligence
Semester A 2018-2019 (Fall 2018)
BGU Computer Science Department
Description of the course
Artificial intelligence (AI) has recently regained the limelight, as the human
world chess champion was beaten by Deep Blue, a program written by a team
of researchers and programmers from IBM. Even more recently, a "re-match"
against a distributed machine in Jerusalem also favored a computer program.
In the more difficult field of partial information and chance games,
such as Poker, AI programs now hold their own against human
champions, as exhibited in a competition held during AAAI-2008.
Two years ago (2016), a computer GO program (Alpha-Go) achieved a major breakthrough by attaining
championship level in game play using deep learning and Monte-Carlo tree search.
In the even more difficult field of automated robotics, a nearly fully autonomous
car has recently (2012) been successful in road tests in the Western
USA (Google driverless car).
Commercialization of the autonomous vehicle is getting nearer as well through the Tesla car,
though there is still a significant path to go until full autonomy and safety is achieved (Tesla-Mobileye
mishap 2016).
Other true AI applications are also
on the rise, from expert systems for diagnosis and advice, through increasingly
intelligent robots, to intelligent and autonomous www agents.
This course deals with the issues of defining intelligence and rationality
in an agent, various methods of formalizing them, and
models for representing and using knowledge. In specific topics, mainly
search, logical reasoning, probabilistic reasoning, and decision-making under uncertainty,
the course will focus all the way down to the
algorithm level, in order to provide some hands-on experience with programming
artificially intelligent agents. We will also briefly examine the currently hot topic of meta-reasoning in search.
Course data and information pointers
- Course Reference: (Introduction to) Artificial Intelligence (202-2-5661: graduate course,
202-1-5151: undergraduate course) - Semester A 2018-2019 (Fall 2018)
- Credits: 4
- Instructor: Prof. Solomon Eyal Shimony
- Graders: Aninoam Yehezkel (yehezavi AT post DOT bgu DOT ac DOT il) - programming assignments, Shahaf Shperberg (theoretical assignments).
- Course scheduled at:
- Monday 16-18, Building 34, Room 14.
- Tuesday 16-18, Building 90, Room 224.
- Syllabus and requirements
- Midterm exams, (use of books and notes allowed):
- 1st midterm exam: Thursday, Nov. 29 at 18:00, Building 26, Room 4.
- 2nd midterm exam: Thursday, Jan. 10 at 18:00, Building 90, Rooms 224, 225, 226.
-
Example midterm from previous years.
- Environment simulator pseducode from Russell and Norvig.
- Some
additional
definitions and algorithms for Bayes networks.
- Belief propagation in chain BNs.
- Additional algorithms for
NN learning.
- Assignments:
-
Assignment 1:
Environments and search (simplified, deterministic). Deadline: Tuesday, November 13 (not strict).
- Assignment 2: Games. Deadline (postponed to): Tuesday, December 4 (not strict).
- Assignment 3:
Theoretical assignment on search and reasoning. Deadline: Tuesday, Nov. 27 (strict, solo submission).Assignment 3 answers.
- Assignment B1:
Logical reasoning (Optional, bonus assignment).
- Assignment 4:
Probabilistic reasoning. Deadline: Tuesday, December 25 (not strict).
- Assignment 5:
Partial observability - decision-making.
- Assignment 6:
Exercise on planning, reasoning and decision making under uncertainty, and learning.Deadline: Tuesday, Jan. 8 (strict, solo submission) (Postponed to Wednesday, Jan 9 at 9AM EXACTLY).
Assignment 6 answers.
- Example quiz and
Answers
- Lecture topics and notes.
Back to BGU CS HomePage