Good news: In-class test on 2/9

Folks

 Another heads-up..  I am planning on having an in-class test on next Thursday. 

I normally only have one midterm and a final, mostly because I don't like to use up class time for exams.

However, since I am any way out of town next week, and since you are all here, it seems we can have 

two mid-term tests (I assume most of you realize that having one more test is better rather than worse for you..)

The test will cover everything we complete until end of this week (but not the make-up class for next Tuesday). 

You will have the opportunity  during Tuesday's class time to come to the class and have a recitation session with the 

TA. 

cheers

Rao

ps: No--I am not trying to convert a missed class into a test--I still fully intend to make-up the Thursday's missed class too by 

an extra meeting+recording during the course of the semester. I am doing the test on Thursday (which is frankly more work for us ;-)

because I have the opportunity to do so without losing class time..

Project 0: Grading & Statistics

Folks,

Project 0 has been graded and will be returned to you in class tomorrow.

The project was graded out of 60 points; 10 for each mandatory part (6 each).  In addition, there were 10 points of extra credit (EC) each for problems 7 and 8, and 5 points of EC for 5(d). The grading for this particular project was quite relaxed, since it was a refresher to Lisp. However, you do need to have at least shown your attempts at solving a problem to have got partial credit.

The undergraduate average was 49.25 and the graduate average score was 56. This does not take EC into account.

If you have any questions about the grading after your projects are returned to you, please e-mail me at krt at asu dot edu.

Thanks.

My absence during the week of Feb 5th

Dear all

As I mentioned earlier,  I will miss the classes on Feb 6th and 8th due to travel. I am hoping to finalize my "make-up" plan by next Tuesday. My current thinking is to make-up/pre-record one of the classes next week, and make the other one up during the course of the semester (probably just before Spring break). 

I am considering holding the make-up class right after the normal class on Thursday (in the same room)--with a short break (so people don't have to come for a separate meeting on Friday). 

This is just a heads up. If you have any comments/concerns, let me know.

Rao

Re: Question on Homework 1

Note that you are searching on a finite depth tree where *all the solutions are at the leaf level, but are distributed non-uniformly. 

This means that if you make an early mistake in choosing the children at the top of the tree, you can get into barren parts of the tree and waste time.

What you need is a way to somehow probe different parts of the final level efficiently.

See if the search algorithm given facilitates this..

Rao

On Fri, Jan 27, 2012 at 12:49 PM, Elliot Drown <edrown@asu.edu> wrote:

Hello,

I am having some difficulty in understanding the algorithm in question 7.2, and why it might provide better performance than other search techniques. Can you clarify the algorithm and the idea behind it for me? Any additional information you can provide would be much appreciated.

Thank you for your time,

Elliot Drown

edrown@asu.edu

Gaming is a hard job, but someone has to do it!

Paper on how popular video games are NP Hard Problems - http://arxiv.org/abs/1201.4995

Next topic--Chap 10: Search with factored or propositional states (rather than atomic ones)

Folks

 For the next topic, we will jump clear across the book and go to 10th chapter--which is titled planning.

It is basically search with propositional states, and makes a very nice follow-on to atomic search we have

done until now. Here are the parts we hope to cover. 

10.1. Definition of Classical Planning ... 366 
      10.1.1. Example: Air cargo transport ... 369 
      10.1.2. Example: The spare tire problem ... 370 
      10.1.3. Example: The blocks world ... 370 
      10.1.4. The complexity of classical planning ... 372 
10.2. Algorithms for Planning as State-Space Search ... 373 
      10.2.1. Forward (progression) state-space search ... 373 
      10.2.2. Backward (regression) relevant-states search ... 374 
      10.2.3. Heuristics for planning ... 376 
10.3. Planning Graphs ... 379 
      10.3.1. Planning graphs for heuristic estimation ... 381 
      10.3.2. The Graphplan algorithm ... 383 
      10.3.3. Termination of Graphplan ... 385 

Rao

Re: [CSE471] Doubt in Project0

This seems to be a question that a lot of you have had.

A clarification for Project 0: If your program seems to give you the correct output, then testing it with a few input cases and showing that output alone will be sufficient for full credit. The output trace is intended for you to show that you tried something if the output is not as expected.

I hope this clears that doubt up. As a reminder, the project is due tomorrow, either at the beginning of class, or in Rao's mailbox before class-time.

On Mon, Jan 23, 2012 at 6:53 PM, abc <xyz> wrote:

Hi,

For the submission of the project it is mentioned to give output trace and stack-trace. I had doubt regarding this, I understand that only if the program does not execute back trace comes up, but for a properly executing program how can I extract the trace.

I tried with the (trace) symbol. But I am not getting any trace output. It will be helpful if it can be explained how to get the output trace and stack trace.

Thanks