Given by Eldar Fischer (Room 625, Tel: 3967).
Teaching assistant: Arie Matsliah.
For a general description of the course and its requirements go to The general course page.
The course: Sundays 14:30-16:30, Taub 5.
Reception hour: Wednesdays at 14:30.
12.4.2007: The final grades have been passed to the secretaries, and should be available in a couple of days.
11.4.2007: Good news - you will be able to pick up Exercise 4 starting tomorrow, and the final course grades will shortly follow.
25.3.2007: The third homework is now graded and ready for the take. The fourth one (and the final course grades) should follow in the near future.
5.3.2007: In the first question in HW4, the graph is also assumed to be connected (otherwise there is more than one stationary distribution and the rest of the question becomes somewhat meaningless).
22.2.2007: The solution to Exercise 4, as well as the not so hard Exercise 4, are now online.
4.2.2007: The deadline for Exercise 3 has been extended to 18.2.
31.1.2007: Some further hints about Exercise 3: The cube question - This one actually uses methods that you already knew when you worked on the second exercise, you don't need to try the more advanced methods. Apart from that, read the hint in the question itself. The "specific subgraph" question - Ok, I'll say it, there is a solution that uses the local lemma. But if you go that way you have to be careful that there are no "almosts" in some of the stages and preconditions. There is a way to do it right. Requests for an extension (a short one) can be heard until Sunday.
28.1.2007: A clarification about the second question: n is the number of vertices of the graph. And a reminder about matingales which you may find useful in solving the exercise: When you define an exposure martingale, it is indeed a martingale (i.e. exhibits the "no memory" property) even if the random object that you expose has a probability space that is not composed of "independently flipped coins". The problem with such spaces is not in the martingale-hood of X0,...,Xm, but in that it is somewhat harder and less automatic to prove bounds on |Xi-Xi-1|.
21.1.2007: The strike is indeed upon us, so there will be no class today. Complementary lectures will be given as a third hour on 28.1 and 4.2, as well as a complementary talk on Wednesday 31.1 at 12:30-14:30 in Taub 5.
17.1.2007: If the strike goes forward, unfortunately it will apply to our course too. Please see the mailing list for the options of getting the lost hours back. Also some good news - the solution to Exercise 2 is now online.
14.1.2007: Exercise 3 is now served. Good luck.
11.1.2007: Complementary lectures are scheduled (in form of a third hour) for 21.1 and 28.1. Hopegully we will not need all of them.
31.12.2006: A very subtle hint on the last question of Exercise 2: To actually do O(log n) steps, you cannot afford to lose all the accomplished work just because of one "lie" in the result of a comparison. Try to make it so a lie will not set you back too far.
24.12.2006: Good news - the submission of Exercise 2 is extended by a week. The new due date is 10.1.
21.12.2006: The official (so to speak) solutions of Exercise 1 are now online. Note that in your graded exercises (come and get them), each question is awarded a score on the scale of 1 to 6.
10.12.2006: Exercise 2 is now served. Good luck.
28.11.2006: Arie Matsliah will be checking your exercises. Please send your solutions directly to him, with the course name and exercise number in the subject.
23.11.2006: There will be no reception hour on 29.11. I will be generally available in other days this week.
12.11.2006: Some clarifications were made in the first question (still the same question, but rephrased a little).
12.11.2006: Exercise 1 is now online. Good luck
12.10.2006: Welcome. As you should notice, from this year there is a separate page for news and exercises, while the "invariant" parts of the course will be in their own page.