CSC421 Intro to AI - Midterm materials
General comments on Midterm
Overall I am satisfied with your performance on the midterm.
I don't think it was particularly hard but I accept the fact
that the time was not enough. In general I graded upwards
frequently even giving 0.5 points for answers that were
totally wrong. Also even if you get perfect scores that doesn't
necessarily mean that your answer is correct. For example
a common mistake in question 3 was to answer 13 models instead
of 3 by thinking that what needed to be proven was that the
sentence is true rather than false. Both answers (3 and 13) were
considered correct if they were justified.
As you can see from the grades several students did
very well on the midterm. These students also have
a lot of courses etc. Also I observed that the students
that regularly attend lectures had much higher grades
on average.
As you can see from the grade histogram most of you
scored around 10/15 with a group of good students
clustering around 14/15 and 15/15. Students with grades less
than 7.5 should seriously reconsider their work in this
course as they are in danger of failing.
Even though in the course outline it states that
students must pass assignments/midterm and final
in order to pass the course I decided to not
enforce this rule for the midterm. I hope this
serves as a wake up call.
I will be going over the questions during
lectures next week when I will also hand
back the exams.
Sections from Textbook:
- 1.1, 1.3, 1.4
- 2.1, 2.2 , 2.4, 2.5
- 3.1, 3.2, 3.3, 3.4 (wout Bidirectional), 3.5, 3.7
- 4.1, 4.2, 4.3, 4.4
- 5.1, 5.2, 5.3, 5.4, 5.5
- 6.1, 6.2, 6.3, 6.4, 6.6
- 7.1, 7.2, 7.3, 7.4, 7.5, 7.7 (wout Circuit-based agents), 7.8
- 8.1, 8.2, 8.3, 8.4, 8.5
- 9.1, 9.2, 9.3 (wout efficient), 9.4 (wout efficient), 9.5 (until
completeness, 9.6
- 10.1, 10.2, 10.3 , 10.5, 10.6
Important Topics
Types of agents
Problem definition (PEAS)
Uninformed search strategies (DFS, BFS and iterative deepening)
Informed search strategies (Greedy, A*)
Admissible heurisitcs - relaxed problems
Local Search (Hill climbing)
Constraint Satisfaction Problems (Backtracking, variable and value
ordering)
Local search for CSP
Minimax algorithm
Alpha-Beat Pruning
Propositional Logic and Inference (Backward Chaining, Forward Chaining,
Resolution)
First-order Logic and Inference (Reduction to PL, unification, backward
chaining,
forward chaining, resolution)
Situation-Calculus (Frame-Effect axioms, Succesor-State axioms)
Representational frame problem