BEN LEONG

Associate Professor
School of Computing

National University of Singapore

15 Computing Drive, COM2 Building, #03-20, S(117418)

Tel: (+65) 6516 4240 Fax: (+65) 6779 4580

Email: benleong at comp.nus.edu.sg

 

Results of CS3243 (Semester 2, 2006/2007) Mid-term Survey + My Responses

The following are the results for the online survey that we did after the mid-term exam. I've also included some of my comments to the mid-term survey.

There are several reasons for this survey:

  • Timely Feedback - We cannot teach well unless we know what's going on and what students do not understand. Waiting till the end of the semester would be too late and there's nothing much that we can do. That said, I hope that students will appreciate that the lecturers are only human and it's not always possible to address all concerns and fix all problems this semester. In any case, the feedback will be taken seriously and future batches of students will benefit.

  • Allow Lecturers to Respond - This survey also presents the lecturers with a opportunity to address the concerns of the class and to clarify some issues from the perspective of the teaching staff, which is also why this particular page exists.

  • Reference for future batches - Students always want to know about class before they take it. The style of the lecturer actually matters more than the subject material being taught. These results will allow future students who are contemplating taking CS3243 with me to know what to expect. If I may cite the key finding in this regard, "Students who are looking for a class that is easy to score, please don't take CS3243 (or actually any class!) with me :-)". My responsibility as a teacher is to help students learn something, not to make a subject easy for them to score. The only way to score in my classes is to learn the material well and to learn CONSISTENTLY. My teaching style is such that students who have a penchant for sleeping during the term and trying to study last minute for the exams will likely do poorly. I apologise for my shortcoming in this regard. Students, please be warned. :-P

Total Number of Respondents : 42/63

Teaching
Question Total Responses
 
Multiple Choice
How do you find the lectures?
A : I have no clue what the lecturer is talking about most of the time 1  (2%)
B : I have no idea what's happening half the time 1  (2%)
C : Lectures are no different from the other classes on campus 11  (26%)
D : Lectures are clear and I am able to follow the material quite well 24  (57%)
E : Lectures are way cool. Easily the best class that I've taken at NUS (not like I've taken a lot, but...) 1  (2%)
F : Others
 
4  (10%)
Skip: 0  (0%)
 
Thanks for the generally positive feedback. The 4 "Others" responses are as follows:
  1. Not really attending
  2. Key ideas are grasped.
  3. Most of time ok. Sometimes can't understand
  4. Lectures are really good. It's one of the best classes i have taken in NUS.


Making lectures available via webcast has a consequence of encouraging some students not to attend lectures. However, I am of the view that the benefits of giving students flexibility and to review lessons that they missed or are not clear about outweighs the cons of having lower lecture attendence. We are dealing with adults and not five-year-old kids. Students should have the flexibility to decide how they want to learn. Whenever possible, I will always try to arrange for webcasts for my classes. :-)

 

Essay
Feedback for Lecturer Ben Leong.
1. Lecture can also go a bit beyond the text book, like exploring other possibilities that are beyond the text book examples given.
2. the speed of speak is a little bit fast, it's better to slow down when teaching important topics. the lecture slides sometime are not quite good. because the background is dark blue, when lecturer highlight some words using red or green, it's very hard to see. overall, he is good.
3. I do not understand why I have been referred to as "champion" by the lecturer. I do not believe that I am a champion, but I do appreciate this act of encouragement. I hope that more of this well-meaning terms will be used to encourage students to show more interest in this module.
4. -
5. You speak clear enough and at a comfortable pace for us.
6. Tried very best to make the student understand the concepts
7. Eh, faster.
8. I think you are really good. Keep it up! I understand that you are just fresh from ur PHD (i think so) and so Sometimes i feel that you know everything but you are not able to bring that out. I think that will improve with time. But apart from that you are great and you have increased my interest in AI. And i think that u are really hard working.
9. Students here are more concerned about grades. So better be careful in the marking scheme.
10. Not bad. But sometimes a bit too draggy.
11. too much Singlish in lecture/tutorial. it's not something bad, at least i m ok with tt personally. the lecture contents, tutorials etc. are all well conducted.
12. The lecturer should talk slower so that students can understand better.
13. Tutorial style is good. Going through key concepts is much better than just doing questions.
15. good lectures!
16. try to be earlier for tutorial.. :)
17. So far, so good =)
18. well done... careful to make sure no one gets lost halfway thru the lec... tries to make lec interesting... helps student to learn instead of throwing chunks of info to memorise
19. Good.
20. Interesting, close to young people
21. Good but sometimes he speaks too fast.
22. Speaking a bit fast.
23. makes the lecture interesting and interactive but maybe not try too hard to remember names (a bit not smooth when replying the questions asked).
Skip: 16  (38%)
 
Thank you for the feedback. I apologize for not being able to go into too much detail beyond the text. This is the first time that I'm teaching this class and I'm barely coping with all the preparation that I need to do each week to prepare for classes. May look effortless, but it's not quite as simple as it looks. Some of you guys should also become teachers in the future and you will find out for yourself. :-P If I however get to teach CS3243 again in the future, I will most certainly be able to do better since I would have worked out all the basic materials. I ask for your kind indulgence.

It's true that I speak somewhat fast at times. Will try my best to slow down -- and I actually I think I do and repeat myself a few times whenever I come to important concepts. Students please do not feel shy about interrupting me and asking questions during lecture if there's something that they miss and do not quite understand. I will not eat anyone. :-) Learning to speak up is important. Those who are confident in public will tend to do better in their future careers.

I already make clear at the beginning of the class that I will have a lot of problems pacing the class. As I mentioned, some will find the lectures too slow and some will them too fast. This is borne out in the comments above. I'm trying to strike a balance and so I think I will maintain the current pace, since going faster or slower is going to upset some people. :-(

Students who think the lectures are too fast can watch the lectures on webcast over again or come to office hours to clarify doubts. Students who think the lectures are too slow can consider skipping lectures and watching them at 2x speed over webcast. Be creative lah! :-)
 

Multiple Choice
How do you find the tutorials?
A : I'm always asleep during tutorials and have no idea what the lecturer is saying. 0  (0%)
B : I have a hard time understanding what is taught in tutorial 1  (2%)
C : Tutorials are okay 11  (26%)
D : Tutorials are clear and helpful in reinforcing the material covered in lectures 26  (62%)
E : Tutorials are brilliant. I understand AI completely because of them. 3  (7%)
Skip: 1  (2%)
 
Thanks for the feedback. Advice for the student who has a hard time following the tutorials : (i) read the lecture notes/textbook and try the problems before you come to class; (ii) ask during class if you don't understand something; and/or (iii) come to office hours.
 
Essay
Please let us have your suggestions on how the teaching for the class can be improved.
1. Can give more examples, and more variations - will help in reinforcing understanding.
2. slow down the talking speed when teaching important ideas. change slides background.
3. Maybe the tutor could focus more on the answer formats we should use to questions. Realised that some questions are quite vague and do not know what type of keyword the paper is asking about.
4. I am longsighted but I am always being forced to sit near the front. This is because the lecturer does not want people to sit behind. It is difficult for me to see the slides on screen, making it hard for me to follow the lecture. I believe that this sort of discrimination is bad for us students on the whole.
5. -
6. It is ok
7. Feel that it is pretty fine.
8. Smaller tutorial classes. This should apply to every tutorial in NUS.. #$%^!
9. You are doing great. As i said the "teaching and explaining" part will become better with time automatically. Just keep going.
10. Examples and scenarios can be prepared beforehand so that there will be less confusion due to improvised teaching in class
11. so far ok.
12. The tutor should be more organised so that every point he wanted to cover should be covered in detail, not just briefly.
13. In my opinion it's quite good now. Lecture scope is a bit wide sometimes, so some stuff cannot be covered in detail. As far as I know, some parts of previous lecture needs to be explained in more detail. (According to feedback of friends)
14. Prepare interesting questions for midterms. Instead of setting a lot of questions for midterm, set a few that will require thinking. In another words, do not emphasize rote learning for mid-term/exams.
16. its interesting, better than the rest of the tutorials ive attended so far
17. Maybe some examples from other Texts instead of repeating whatever have been mentioned in AIMA?
18. nil. =)
19. hmm not really sure... more animations? hee
20. Lectures are ok.
21. Cannot expect everyone in tutorial class understand some concept. Best if can go through a bit before go into the tutorial question.
22. maybe can have a variety of questions that is able to cover all aspects in terms of examinable areas. sometimes, i feel that it is not the understanding but rather the presentation of answers. therefore, without the coverage, I might not even know what the question expect.
Skip: 20  (48%)
 
More examples? More variety of questions? I guess these students are somewhat perturbed by the seemingly wide scope in the possible questions that can appear in the exams. Unfortunately, this is the way the class is. Students have to get used to the idea by now that some subjects are not like the 'A' or 'O' Levels and it's not possible to have a 10-year series.

I did provide the class with the past year midterms for practice before the midterm. Students can get the past year final exams from NUS Library. There are also a lot of practice questions in the text. Furthermore, students can also surf the web and try out the AI exam papers/homework from other universities. AIMA is a text that is commonly used by MANY schools.

I will try to speak a little slower and I will try to improve the colour scheme for the slides to make them clearer. Long sighted student should have told me that he/she cannot see! Discrimination is a very serious accusation that should not be thrown around lightly! :-) Will left students sit wherever they want during lecture. :-P

Smaller Tutorial classes is not possible unless the school assigns me more teaching support. I am handling everything from lectures to tutorials to grading for this class of 63 students. Yes, please complain at the final survey that will be conducted by the School at the end of the semester about tutorial class size. Maybe then I'd get TA support in future semesters. :-)

W.r.t. comment on examples, I believe the student is referring to tutorials. I actually have my examples in mind before the class starts, but sometimes students say something smart and it occurs to me on the spot that there might be a better way to explain something so I fumble a bit. Apologies. Usually happens only for the first tutorial. By the third tutorial, I would have ironed out all the bugs. :-)
 

Homework
Question Total Responses
 
Essay
There's very little "homework" for this class, but there are tutorial questions that students are supposed to prepare for each week. Please let us have your views on the tutorials and/or suggestions for improvement.
1. It would be great if we had more problems to practice on. Not that I am requesting for more homework but references to materials that contain questions for practice etc.
2. I prefer to have more questions to practice more.
3. Tutorial questions are quite workable and helpful.
4. The tutorials are sufficient.
5. -
6. Tutorial question are fine. But I do have difficulties in doing proofs. Maybe it has always been my problem as I may understand a concept but not able to write it down in words to explain what I think.
7. Tutorials further enforced the concepts
8. More theoretical?
9. they are just fine...may be increase the number of questions so that we can have more practice. Like i felt that all the topics are not covered by the questions in the tutorials. For example i felt there could have been more questions on CSPs, local search etc
10. Tutorials are fine. Some unclear concepts are explained better in tutorials
11. I'm fine with how tutorial is being carried out.
12. so far ok.
13. The tutorials are good in that they require students to think and understand the concepts. However, it would be the best if the lecturer could at least have drawn some parallels between the discussion in lectures and tutorial questions so as to aid in the understanding.
14. Interesting questions that helped me to understand key concepts. Good.
15. Tutorials are fine.
17. tutorials are to the point and specific
18. The tutorial questions are well chosen, and the method of using the questions to aid in explaining concepts rather than merely going through answers is good.
19. tuts are gd... qns planned in a way to reinforce understanding of concepts... lecturer knows what he is doing...
20. They're fine.
21. No comments.
22. The tutorial questions are okay.
23. maybe can have a mini online mcq quiz (redo-able) to aid students to understand the concepts, which will generate the answers right after, so that students can have a check on their own understand.
Skip: 18  (43%)
 
Re: questions on more practice questions, I will quote response from earlier question, "There are also a lot of practice questions in the text. Furthermore, students can also surf the web and try out the AI exam papers/homework from other universities. AIMA is a text that is commonly used by MANY schools".

As for proofs, that's not something that we are supposed to be covering in CS3243. That's the topic of CS1231. Students may wish to review their CS1231 notes. Mini-MCQ quiz is a great idea.... unfortunately, I don't have the resources to put up something like that. Will keep this idea in mind. Maybe someday we can work something out.
 

Essay
Please let us have your views on Mini-Project: Breakthrough
1. Please put up a test webpage so that it actually tests some cases ( sample cases) on the *sunfire* computer, so that we know how well our program performs on the sunfire computer also.
2. It was fun and simple..
3. I feel interesting and learn a lot.
4. In a way unfair as students who already have AI background obviously will fare much better than beginners who are using this course as an introduction.
5. The game is too abstract for me to understand. I can barely figure out how to play the game, let alone try to win. All projects based on games need to be easier to understand, and be simple enough so that it is possible for weaker students like me to even visualise a path for victory. Even after the mini-project, I am still not sure what makes a winning position for Breakthrough.
6. Could have been fun...
7. It was a interesting and fun project. Was hoping that we could see some of the techniques used by the top few students to help us learn.
8. Pretty helpful in understanding some important concepts and also an initial idea of how AI is written
9. Fun! Could have a pre-match halfway through to let people test out agents.
10. it was fun to code.
11. I think more considerations can be made to Non-SOC students taking this module. Such as Sunfire accounts and java programming.
12. ARGHHHH!!!!!!!!!
13. it's an interesting assignment.
14. The project was fun as I had pushed both myself and the agent to the limits when trying to make things work better.
15. Fun project. Not very hard to code too, except that the competitive element might be a bit harsh. If some group can't even finish the minimax they can't really implement any form of decent evaluation function.
16. Good introduction to AI. Though might be better if students are allowed to choose their own project (like the final project). Reason being that this will allow students to focus on what they want to learn instead of just learning alpha-beta, minimax, evaluation function.
18. breakthrough jumps us into the problem directly which disorientated me. although I know about the general theory and algorithm I have no clue about how to apply them into the program. the algorithms taught are very general and hence application is made difficult
19. Too much time spent on this! The AB Routine was easy and fair enough.. The optimization was time-consuming.. Maybe more percentage of marks could be given on this such that the time spent is well-justified.
20. It's interesting and eye-opening, and I had fun.
21. fun... but a bit painful coz started in wrong direction =P
22. It was ok.
23. Interesting.
24. It's cool
Skip: 16  (38%)
 
I'm not sure that I understand point 1. What samples are required? Students can always log into sunfire to see how fast their programs run. Suggestion of mid-way prematch is not practical. Most teams were doing their projects near the deadline! :-) I know that some students had mini-matches among themselves, so this can be done among yourselves. I have no objections to students playing their agents against each other prior to the submission, as long as code is not being shared! :-P

I'm glad most students enjoyed the mini-project. Let me get this clear: there is a GOOD reason for the Mini-Project and that's to MAKE the students learn alpha-beta. :-) For this reason, I cannot allow a free-for-all project and I cannot have the tank project before Breakthrough. Quite sure that all those who spent effort on the Mini-Project learnt their alpha-beta.  :-)

I cannot release the source code by the better teams. I cannot release source code period since this assignment is likely to be used again in future semesters. In any case, the learning value of the assignment is in the coding of MINIMAX and alpha-beta and the process of exploring the evaluation functions. Learning how to play Breakthrough well is pretty pointless.

Re: point about consideration for non-SoC students, I don't buy that non-SoC students have been disadvantaged in any way. There was a note in the FAQ on getting accounts on sunfire. Students who don't have access to sunfire should have informed me and I would have dealt with it. Students are also allowed to use C++ instead of Java to complete the assignment (and many did do so). It's true that those who use Java has some minor advantage because the GUI is only available in Java, but given that I have limited resources for this class, it is clearly better to provide tools that benefit a greater number of students. One student has been kind enough to contribute his C++ game driver. Students who want to use C++ in future semesters will therefore probably be better off.
 

Multiple Choice
On average, how much time do you spend doing homework for this class per week? Take the total time spent preparing for Tutorials + doing Breakthrough and divide it over the 6 weeks prior to the mid-term break.
A : At most 10 hours 23  (55%)
B : Between 11 and 13 hours 8  (19%)
C : Between 14 and 15 hours 6  (14%)
D : Between 16 and 18 hours 0  (0%)
E : At least 19 hours 2  (5%)
Skip: 3  (7%)
 
Seems like the majority of students don't spend a lot of time on this class. I'm quite shocked to see that two student have spent more than 19 hours per week on this class. I would be curious to understand why they are spending so much time. Perhaps the two of them can send me anonymous feedback via IVLE to help me understand their situations.
 
Essay
What are your views about the Final Project?
1. I am facing problems understanding the code for tank game. Also the documentation an be a bit more detailed.
2. Difficult and interesting
3. I think it should be fun and challenging
4. Hope it'll be interesting.
5. This is much more easier to visualise compared to Breakthrough. At least I can relate to this game. I am confident that I will be able to do something for this final project, compared to Breakthrough. Perhaps, Breakthrough should have been the final project, and the tank project the mini project. This would interest me in the module a lot more.
6. Can be interesting and fun... Good idea to have open projects.
7. Interesting and challenging. The research part of the project really made me go online to find techniques used by others to solve similar problem (Tank Game)
8. Provide a chance to further illustrate more components about writing AI.
9. Out of ideas. Some advice on possible projects would be nice.
10. it's ok.. just that i wanted to work with lego sets which i couldn't.
11. Potentially interesting and challenging
12. Too broad that it's scary.. Maybe could define the scope a little.. Like maybe limit it to games, scheduling or something like that..
13. I'm still not very clear about the Tank game APIs
14. The final project is a good way to help students learn by themselves. However, for the tank game, more details should be provided because the java documentation may be insufficient.
15. Finding a suitable problem was kind of tricky because we don't really know what actually qualifies. But working on a self declared topic will be fun... until we find that it's too hard to implement. (hopefully not)
16. Great idea. Better to do project then to sit for exams. Good to see there is not much emphasis on exam for this module!!
17. I would like to suggest to have a higher weightage for the final project should constitute more marks in CA. Final project emphasizes more on the documentation than a functional AI agent.
18. its an interesting project. However the project lacks a debugging feature (the source codes are locked hence we cannot use our preferred IDE to code the programs.
19. Too many factors to consider!! What is the aim of the Final Project? to derive Strategies? Knowledge Base Agent? Logic? What realm of AI is it suppose to be in?
20. Interesting and fun. I really appreciate the effort to let us choose what we want to do, and all the help and clarification that has been provided through IVLE and the lectures.
21. tank game project will be fun when completed... looks like a lot of maths to figure out (-.-') kudos to lecturer for coming up with the environment...
22. Looks interesting.
23. No comments.
24. Time consuming, quite challenging anyway
25. The idea of able to do some researching is quite good, expose us to the uncertainty and more resources. but afraid that there might not be enough time as we also have other ongoing assignments. In regards to the tank game, it seemed interesting to be able to compete with other teams (just like breakthrough).
Skip: 16  (38%)
 
There is nothing much I can say to the above comments. As for the Tank Game documentation not being completely trivial. Well, that's life. Part of the Tank Game project is supposed to make you guys struggle with a pre-compiled library. Debugging that library is my problem and not yours. You job is to debug your own agents. You can use any IDE you like, just configure cs3243.jar as a library in the IDE's project.
 
Midterm
Question Total Responses
 
Multiple Choice
Comment on the difficulty of the midterm. On whether the questions are easy or hard, not the length of the midterm (we know the midterm was a little too long). Apologies. :-P
A : Way to easy 0  (0%)
B : Somewhat easy 3  (7%)
C : Just right 20  (48%)
D : Somewhat hard 13  (31%)
E : Way over your head 3  (7%)
Skip: 3  (7%)
 
 
Multiple Choice
Do you think the midterm is fair? Again, factor out the time component. The question here is whether students feel that the scope of the midterm fairly reflects what was covered in class, i.e. that the questions asked reflect what students should realistically be expected to know.
A : Yes 35  (83%)
B : No 3  (7%)
Skip: 4  (10%)
 

Glad to know that most students think that questions are fair. Be prepared to see more of the same at the Finals.
 

Essay
Any other comments/suggestions for the mid-term?
1. Shorter paper please :(
2. i think a couple of multiple choice Q would have been better.
3. Make it shorter
4. no.
5. Questions sound a bit vague, dunno how to answer the questions, this is with regards to the manner of approach to answer.
6. The last question in the paper involved a big mathematical concept. I was unable to understand the question. More pictures would have helped. I also believe that more measures should have been taken to enforce test rules. Some people took extra time (time before and after the test started) to continue writing down answers. I hope that in the future these students will be punished for not following test rules.
7. -
8. I have problems with writing proofs thus do not know how to put in my points. I may understand the material taught in general but writing proofs is somewhat difficult for me.
9. Suppose is still the time issue, as there is no time to carefully understand the question and to think through the answer. Suppose not everyone can think that fast
10. i think it was good. good questions. That is how a test should be. Please keep the same difficulty level for the finals.
11. Need more time.
12. I think the time part is fine :)
13. i like this format. hate MCQs esp when the options are similar and confusing.
14. I discovered that even though I seem to understand what is taught somehow I cant draw a parallel to the mid-term questions. Hence, it could be best if the lecturer give more examples each time so that the topic is thoroughly understood
15. Nothing much. Testing based on understanding is good. (in the long run)
16. Too many questions.
18. time. probably.
19. Q1 is too much of a giveaway.. Q2 requires the domain knowledge that to give disc from one peg to another peg requires 2^d -1 moves.. Not good.. Q3 doesn't test much on AB but more like a puzzle to solve.. like IQ Questions.. Q4 is hard despite the lecturer saying giveaway marks.. :(
20. It was frustrating that I studied yet my results don't reflect that.
21. was a bit blur when i reached the last qn abt the maths equations... under tight time constraint couldn't really understand qn at first... but understood in the end... is it reasonable to suggest less mathematical qns for exams as it takes some time to understand maths probs with explanation in words?
22. Nope.
23. Too many questions and time given too short.
24. Just keep the same format for the final, although grades are not good as expected, i still think the questions are quite formed and reflect the ability of student.
25. Some questions didn't actually test whether students really understand or not, i.e., they require a bit of "familiarity" with the problems
26. A bit long for me, but it's ok because this can be used to determine the response of students. I realised that I really do not understand consistency and admissibility.. haha I like the nim game sort of question, kinda test what we learned in class, the minmax, the state reduction thingy.
Skip: 15  (36%)
 
Final exam will be shorter (relative to the available time). Personally, I don't like MCQs, 'cos some can get the answers right by guessing. Aiyah, when I said that there's giveaway in Q4, I don't mean that the entire question is a giveaway. I thought the question to draw the constraint graph was a giveaway... but apparently it wasn't quite. :-(
 
General
Question Total Responses
 
Multiple Choice
What is your overall impression of the class thus far?
A : This is a horrible class. Truly regret choosing it. 1  (2%)
B : Not so great compared to my other classes 5  (12%)
C : Just like any other 13  (31%)
D : Better than my other classes 16  (38%)
E : Definitely one of the best classes that I have taken in my life. 3  (7%)
Skip: 4  (10%)
 
My sympathies to the poor student who regrets choosing CS3243. Perhaps he/she can make an appointment to see me and we can see how we can help him/her out?
 
Essay
Would you recommend the class to other students? Why or why not?
1. Would recommend for the content of the courses. But would not recommend if he/she is grade oriented.
2. I would recommend this to my friends.. its more application-based and just pure theory and also like the higher weightage of CA instead of main exam.
3. Yes. I think I can learn sth from it
4. I don't know but if someone ask me if this course is easy or hard to get marks, I'll say it's not hard to get high marks but not easy either.
5. Interesting course, I will recommend it to others.
6. I would recommend this class to students who know how to play chess or understand how to play Breakthrough. Otherwise, abstract projects like Breakthrough will likely traumatise them for life. I know that I have been unable to shake off dreaming about a 6x6 grid every night.
7. Yes if they are motivated, hardworking and serious about their studies. If they are only here to dream, then no.
8. Yes if they are interested in AI and similar fields. No if they are not interested or are not too good with proofs.
9. I would recommend the class as the concepts learnt are interesting. However, i would also mention that it is not easy to grab some of the ideas.
10. I'll ask them to throw a coin.
11. Yes of course. it's a lot of fun
12. Not really. It is not an easy module to score
13. Yep, 'cos the lecturer is better than the other lecturers :) though the content is rather "informative"
14. yes. the topics covered are quite interesting. the algorithms presented are amazing.
15. I would recommend the class to students who have the passion to learn AI.
16. Yes. AI is interesting. (at least to me)
17. Yes. I like AI.
19. yes. the other students have expressed interest in taking the course but have no idea of how tough/difficult the course is because while the interest is there, the cap score is more important for the time being (for them) and they have to "check out " the mod before they decide
20. Yes, if they have an interest in programming and/or artificial intelligence, this class is a must-take.
21. yes... interesting and fun course... lecturer puts in effort to make sure students really learn something...
22. Yes, I'd just warn them that there's a lot of searching involved. Lecturer is good.
23. Yes. Interesting.
24. Yes.
25. yes! learnt quite a lot, in terms of how to make computer understand and think the way we do. maybe quite useful in future if we were to develop games or some other application. very interesting.
Skip: 16  (38%)
 
Honestly, this grade-oriented thing irritates me quite a bit. After being at NUS for about three years, students should have figured out by now that grades are assigned on a curve. Therefore, some will get A's and some will get B's and C's. While how easy it is to get a good grade in a class depends to some extent on who else is taking the class, most of it depends on how much effort a student puts into the class. Easy to score means what? Easy to get a good grade without putting in effort right? If so, it probably means that you're not required to be learning anything to begin with, which begs the question, "Why are you taking this class?" I do not teach mickey-mouse classes. Those who are looking for easy to score classes without putting in effort, please avoid the classes that I teach. My apologies to the poor students who didn't know this before choosing this class. Hopefully, with these survey result, future grade-oriented students will know better. :-)
 
Essay
Do you have any concerns about the class? If so, what are they and how can we help?
1. None.
2. no.
3. Yeah, fared much worse than I expected to. I love the contents of the course, but did quite badly for the project and mid term. I guess I didn't realise the focus of the answer format used in the term tests.
4. Yes, I am very concerned about the class. I hope that the class will not be cancelled in the future. I hope that there will be more tutorials so that I can understand the lessons better. Perhaps more questions can be introduced next time so that I can practice more.
5. -
6. Sometimes it is not that I do not want to ask question but I do not know what to ask. And I am quite bad at writing proofs i.e. question that say show that "this correct" or "this is X". I understand that this is the basic that is taught in other courses but I just could not properly write them down. Or is there a way to show my answer without writing formal proofs like drawing diagrams etc?
7. No major concerns. Just hope to learn something from the class and obtain a acceptable grade.
8. none actually
9. Sometimes students do not understand but shy to ask. Most of the time actually.
10. No major concerns :)
11. so far ok
12. Yes. The lecturer needs to draw parallels between the things taught in lecture to some problems. Even though some of this has been done, most of it is done briefly in tutorials. It would be best if the lecturer explains how to understand the topic better -> slowly.
13. No.
14. The classes are fine :)
16. no concerns
17. nil. =)
18. a bit overwhelmed by lots of algorithm... able to understand everyone generally but sometimes not 100% sure of what 1 or 2 lines in the algo mean... any advice? =)
19. no
20. No
21. nope
Skip: 21  (50%)
 
CS3243 will never be cancelled. AI is a fundamental course that will be taught every semester. I did try to ask the Department to cancel CS3243 this semester so that I can teach a networking class instead, but they didn't agree.

Students who didn't do quite so well on the midterms, please don't worry too much. The midterms aren't worth all that much. If you put in the effort, it is possible to make up for it in the Finals. The Finals will be of the same standard as the midterms, so people who really know their stuff will be able to demonstrate convincingly that they are better than the other students, even if they had messed up on the midterms.

Re: point 6. There are many ways to prove things. Some things are proven with equations. Others can be proven with diagrams. There is no general answer to your question. It all depends on the actual statement that you are trying to prove.

As for the "I got question, but I'm shy to ask" problem, I don't really have a cure. Those who are truly shy can consider sending me anonymous emails via IVLE? I guess if it's anon, then I can't respond directly, so likely what I can do is to reply on the IVLE Forum or something. Generally, it will do students good to overcome their fear of public speaking. In life, those who can talk (or dare to talk) have a distinct advantage over the shy folks. You can either choose to learn how to speak up while you are in school, or you can choose to suffer when you go out to work. Your choice. Pick your poison.

Re: point 18 on understanding algorithms. Actually, understanding the key ideas is most important. To understand complicated algorithms, the easiest way is to simulate the algorithm. Come up with a simple example and run the algorithm on it. See how the data structures change and it should be clear after one or two examples, what's going on.
 

Essay
Final question: tell us more about anything else you have to say about the class that is not already covered by the previous questions. Phew. :-)
1. None.
2. As lecturer said, I come here to learn something. Hopefully, I can reach the goal at the end of semester. ^_^
3. This course has a lot of people who really know stuff about AI beyond what this course is teaching. It is disadvantageous for the students who are totally new to this course.
4. The guest lecturer, Mr. Martin Henz, was very intimidating. I do not understand what he was talking about. I feel inadequate when he was asking questions about his topic. Possibly this is due to the fact that I was still stunned by the midterm test. Perhaps it would have been better if his talk went before the mid-term test.
5. Not in a position to comment due to lacklustre performance and poor attendance. Hard to follow class if not motivated.
6. I think in general the material covered till are quite good and useful. The project really helped me understand how alpha beta and minimax works. No complains.
7. Basically, the class is interesting and i managed to take away with me some interesting AI concepts that would allow me to explore further into the area.
8. More in depth?
9. phew... the survey was long :-) a nice gesture from the lecture.. i hope the feedback will actually be read. i have always wonder whether lecturers actually read the surveys or not. So if you are reading this.. you will post a difficult question on some AI topic that we have covered in the forum with the topic - "Something for You guys to think about"
10. To improve, either we adapt to lecturer's style or lecturer learns our way.
11. everything is covered ...
12. Nothing much. Just slow down a bit and the lecturer will be fine.
13. None. =p
14. It is an introductory module on AI. Can't really expect much I guess :-)
16. interesting course. Should there be tutorial questions modeled after the mid term examinations?
17. Which other classes do you teach? =)
18. keep up the gd work doc! dun be too concerned with malicious comments! dun think they are based on solid arguments
19. no
20. Hopefully, only key ideas and understanding will be tested in exam. The reason is that the lecture didn't actually go in to much depth (and that's fine as those stuffs are mainly theoretical, i.e., too general) while the projects have been taken a considerable amount of time.
21. nope
Skip: 22  (52%)
 

Re: claims that there are students who already know AI and therefore it's disadvantageous to the other students. One, I have no idea why students who already know AI are taking this class. Two, so what if they know some things already? I believe that it is possible for those who come to class not knowing a thing to catch up and be even better if they put in enough effort. Quit complaining lah. It's not going to change things and there's nothing the lecturer can do about it. Cannot possibly ask these folks, even if I can identify them, to erase their memories right?  If these students know more, it only means that they had spent the time and effort earlier learning the stuff. Good for them. You do the same, you'd be as good.

Sorry about the poor student who got intimated by Martin. I must admit I was pretty shocked myself when I saw him prancing among the audience like a talk show host. :-) Consolation: material on guest lecture will not appear on the Finals. That lecture was purely for your edification. I do not however renounce the right to set a question on applying CSPs to game scheduling.... as long as anyone who did not attend the guest lecture can be reasonably expected to solve the problem. :-P

Tutorial question modelled after midterm questions? You will see some tutorial questions in the future weeks that are past year Final exam questions. Fair enough? Again, I must emphasize that it is not possible to provide you with the entire of possible questions that can appear on the exams. Tutorial questions are chosen to illustrate specific teaching points for the class. Also, I only have one hour of tutorial a week and so I have to be judicious about what exactly is important and worth emphasizing. Students who want more practice questions can find them in the text and also on the websites of other schools that also use AIMA as the text.

Re: point 20. You bet I will try to set questions on the Final Exam that are pure application questions such that memorization cannot possibly help. People have to understand the concepts before they can solve the questions. Truthfully, the concepts that people need to understand for CS3243 are quite simple and shallow (I did tell you guys this at the first lecture did I not?). Nevertheless, they are very fundamental and they are the basis for more advance AI classes and for other Computer Science classes.

What classes do I teach? For now, the only class that is confirmed my baby for the next two or three years is CS1101S (Scheme). I suspect that this will be the first and last time that I will be teaching CS3243 for a while (but one never knows). My area of research is networking and distributed systems and it is very likely that I will be teaching CS2105 and CS3103 at some point in the future. I think I'm teaching CS1101S next semester (2007/2008 Semester 1). :-P
 

 

 

Last updated $Date: 2015/02/26 13:12:27 $