Visa resultat

Stochastic optimization algorithms, FFR105

Teaching

1. How do you rate the lectures?

Not good at all»		0		0%
Not so good»		0		0%
Neither good nor bad»		1		2%
Good»		12		35%
Very good»		21		61%

Genomsnitt: 4.58

- I don"t like that not all lecture time was used. It would have given much more if it were. If there is no more material to go through then it would be good to recap something more, or do more exercises. It is really good to see things more than once, and if there anyway is time left then I think it should be used.» (Neither good nor bad)
- Certain (mathematical and proof) lectures felt a bit rushed.» (Good)
- I did only attend 3 lectures though, due to lack of time in my own schedule. » (Good)
- I didn"t feel sleepy during the Monday morning lectures! (usually I do during morning lectures) » (Very good)
- Very well structured and thought through. Easy to follow.» (Very good)
- Very structured! Mattias seemed a bit stressed sometimes, although we never ran out of time.» (Very good)
- I especially liked the mixture between slides and blackboard. I"m impressed by how swift and problem free the change between these were. » (Very good)
- Wahde is the best lecturer I"ve had at Chalmers. The lectures are easy to understand with clear presentation leaving no question-marks.» (Very good)
- Very clear and thorough.» (Very good)

2. To what extent have the lectures been of help for your learning?

Not at all»		0		0%
Some extent»		6		17%
Large extent»		16		47%
Very large extent»		12		35%

Genomsnitt: 3.17

- knew a lot beforehand» (Some extent)
- Interesting lectures, not much information that couldn"t be found in the book (not necessarily a bad thing). Good examples and videos for usage of algorithms.» (Some extent)
- Most understanding for learning is done when doing the matlab code.» (Large extent)
- I love lectures.» (Very large extent)
- See above.» (Very large extent)

3. Do you think that the lectures complemented the course literature well?

No, not at all»		1		2%
Not so much»		3		8%
Yes»		14		41%
Yes, very much»		16		47%

Genomsnitt: 3.32

- The lectures were the course literature, so there"s really no complementing as such, however it is easy for the student to have written material that more or less exactly corresponds to the lectures. Easy to check things, plus you get necessary repetition automatically.» (No, not at all)
- They were exactly the same. It would have been nice with some examples not in the book. And also maybe some paper by another author to get another view of the subject as well.» (Not so much)
- As usually when the lecurer haw written the book, the book tend to be a bit like lecture notes. Sometimes, even the words are exactly the same. But I don"t think there"s much to do about it.» (Not so much)
- Did only attend three lectures though..» (Not so much)
- They pretty much said the same, but I like to read about what I"ve heard so I liked it.» (Yes)
- I think that the course literature covered the essential part of the course in a very good way. What the lectures brought in addition to lectures was inspiration and an overview of the applications.» (Yes)
- Lectures provided good information around the theory in the book.» (Yes)
- I have not read the literature extensively, but the parts I have read were made clearer with the lectures.» (Yes)
- While the literature is clear in itself, the lectures were good at pointing out important things.» (Yes, very much)

4. To what extent has the Matlab introduction been of help for your learning?

Not at all»		4		12%
Some extent»		17		51%
Large extent»		4		12%
Very large extent»		8		24%

Genomsnitt: 2.48

- I did not attend the lecture but the assignment was good.» (?)
- It was too basic. It felt like the prerequisites assumption was that the student had never seen matlab before. Also, to copy (by writing yourself) the code written on a paper does not give that much understanding. It would have been better to describe what should be done and give some hints and let the student come up with the code. » (Not at all)
- Didn"t really do it. But for those who dont know Matlab I can see that it is very good. Also good for setting the coding standard.» (Not at all)
- Rubbing in the coding standard was actually good, even if I thought it was a waste of time while doing it.» (Some extent)
- Not with actual Matlab syntax but with the coding standard.» (Some extent)
- It probably would have been more beneficial to do some programming from scratch instead of copying code from a paper.» (Some extent)
- The material wasn"t really new to me, just the implementation.» (Some extent)
- Not very much though, but I allready knew matlab quite well. However, it was nice to have something to go on for the first home problem.» (Some extent)
- I had previous matlab experience, but it was good to get to know the wanted syntax.» (Some extent)
- It was useful for the first home problem, but the most essential was to have a set of code to work from since GAs contain many elements.» (Large extent)
- Very good introduction. It served three purposes at once: teach coding standard, develop matlab programming technique and teach how the GA works.» (Very large extent)
- Since I took a year off between Bachelor and Master I had forgotten most that I had learned so this was a very good review, plus you learned some new good tips of how to implement a matlab program.» (Very large extent)
- It could have been better if we had some other classes for other EAs. The Matlab introduction session was so useful and could help me alot to understand the GA and how to use it.» (Very large extent)
- Have not used matlab in 2 years» (Very large extent)
- It was very good to have a sound basis on which to develop the programs for the home problems from. It could otherwise have been tricky and messy.» (Very large extent)
- It was a good way to get going.» (Very large extent)

5. To what extent have the home problems been of help for your learning?

Small extent»		0		0%
Some extent»		2		5%
Large extent»		6		17%
Very large extent»		26		76%

Genomsnitt: 3.7

- It is a practical course, so the theory taught need to be practiced in order to learn it.» (Large extent)
- Problem 2.3 was more time consuming than 2.1 yet 2.1 generated more points. Maybe provide a pre-coded model for the truck. It"s hard to plan when you can"t use the points as a measure for time required to solve.» (Large extent)
- The only point is about grading and scoring. It is too strict. I think everyone by studying the Home Problems can get if the student has learnt the main subject or not. Deducting points based on some small points could distratc the main aim of learning.» (Large extent)
- Indeed very good home problems.» (Very large extent)
- Really good to do everything in practice!» (Very large extent)
- This is the whole point of the course. I would actually have liked to have more home problems, but a little more evenly distibuted throughout the course. » (Very large extent)
- I think home problems are generally a very good way to learn and the problems in this course was really good. I think however, that they would have been even greater if one had been allowed to use a faster programming language than matlab, for example c or c++. Some of the assignments in HP2 required a lot of computing time and I do not think that the point of the course is to sit and wait while the slow matlab code executes. It might encourage people to think harder in order to avoid a lot of computing time, but I still think matlab is a bit to slow to run some of the algorithms successfully.» (Very large extent)
- Interesting, really great!» (Very large extent)
- Studying a home problem gives much more in return rather than studying exam questions.» (Very large extent)

6. What is your impression of the course literature?

34 svarande

Very poor»		0		0%
Poor»		0		0%
Average»		4		11%
Good»		16		47%
Very good»		14		41%

Genomsnitt: 4.29

- The price of the book keeps the grade down unfortunately» (Average)
- There"s nothing about stochastic optimization with equality and inequality constraints...» (Good)
- Since it is the only material I have seen in this subject it is hard to compare it.» (Good)
- Possibly a bit short. But perfectly balanced in terms of examples, theory etc.» (Good)
- The book is good, just not very different from the lecures.» (Good)
- Although a bit expensive.» (Good)
- Very well written book, liked it a lot! A bit too expensive though, but I know the reason.» (Very good)
- This book is very good and to the points, but in some parts there are needs to more explanations and sometimes need to solve more and calculational examples.» (Very good)
- Clear and easy to read, but a bit limited in scope. Personally, I would like a more detailed, comprehensive overview of the field. » (Very good)
- Very good! Contains all you need for the course, perhaps its a bit too much so. No extra theory if you would like too read more.» (Very good)
- Simple, clear and to the point.» (Very good)
- It covers the course material very well, i.e. all relevant material can be read in the book.» (Very good)

7. How well did the course administration, web page, communication etc. work?

Very badly»		0		0%
Rather badly»		0		0%
Rather well»		9		26%
Very well»		25		73%

Genomsnitt: 3.73

- Sometimes, lecture files uploaded on the webpage one week after the lecture.» (Very well)
- Excellent information.» (Very well)
- Very good information. There was never any informational doubt on anything.» (Very well)
- Everything was where it should be. No complaints.» (Very well)
- Compared to other chalemrs homepages, realy good. Greate structure aso. » (Very well)
- Information was clear and mostly up to date.» (Very well)

Your own efforts

8. How many hours per week did you spend on this course?

34 svarande

At most 15 hours»		3		8%
About 20 hours»		21		61%
About 25 hours»		5		14%
At least 30 hours»		5		14%

Genomsnitt: 2.35

- Almost day and night during the last week of the second home problem. Portioning out the homeproblems into smaller pieces would be preferable.» (At most 15 hours)
- towords 25» (About 20 hours)
- A lot more when working on home problems, and more like 10 at other times.» (About 20 hours)
- On average about 20 hours, but much more during the week before the handin of home problems 2.» (About 20 hours)
- Sadly the ANN course took much more time than it should have, making this course suffer.» (About 20 hours)
- More when home problems were due.» (About 20 hours)
- Including the exam preparation (derive proofes etc.). The second set of home problems was rather time consuming.» (At least 30 hours)

9. How large part of the teaching offered (lectures and exercise classes) did you attend?

34 svarande

Less than 20%»		2		5%
20-40%»		0		0%
41-60%»		3		8%
61-80%»		6		17%
More than 80%»		23		67%

Genomsnitt: 4.41

- Missed one week due to external work, been to most of the rest.» (61-80%)
- Have all lectures in the morning or none of them. » (More than 80%)

Study climate

10. How were the opportunities for asking questions and getting help during the course?

Very poor»		0		0%
Rather poor»		0		0%
Rather good»		5		14%
Very good»		23		67%
I did not seek help»		6		17%

Genomsnitt: 4.02

- Lecturer replied to my email fast.» (Very good)
- Thanks for answering e-mails so quickly.» (Very good)

11. How was the course workload?

Too low»		0		0%
Low»		2		5%
Adequate»		20		58%
High»		12		35%
Too high»		0		0%

Genomsnitt: 3.29

- Adequate, but I didn"t take the neural network course.» (Adequate)
- Good workload, perhaps a bit more time for HP2 would have been nice and less for HP1. The second one was more time-consuming and much more interesting! (Though if you only do the problems for grade 3 I guess the time is plentiful).» (Adequate)
- It"s high but I think it should be that way.» (High)
- It was fine since it was interesting. In other courses I wouldn"t have liked working till 3 am, but in this one it was kind of nice.» (High)
- Combining this course with Neural Netorks made the time of doing the homeproblems rather short when the deadlines coincided.» (High)
- Low in the beginning and high in the end. Maybe the workload could be distributed more evenly over the period? Maybe the home problems could be divided into three, introducing the first one a little bit earlier?» (High)
- It was quite high, especially to get above the required points on the exercises. This due to other courses also being of high workload.» (High)

Course goals, level of difficulty, exam

12. How understandable are the course goals?

34 svarande

I have not read the goals»		7		20%
The goals are difficult to understand»		0		0%
The goals give some guidance, but could be clearer»		6		17%
The goals clearly describe what I am supposed to learn»		21		61%

Genomsnitt: 3.2

13. Is the level of difficulty of the course reasonable, considering your background and the number of credits?

No, the level is too low»		3		8%
Yes, the level is reasonable»		30		88%
No, the level is too high»		1		2%

Genomsnitt: 1.94

- Slightly too low.» (No, the level is too low)
- It"s quite a big workload, but that hasn"t anything to do with the difficulty.» (Yes, the level is reasonable)
- A bit hard on the mathpart compared to prerequisites "basic engineering mathematics"» (Yes, the level is reasonable)
- A bit on the low side. I am not a high results achiever normally, but i think this course went pretty well.» (Yes, the level is reasonable)
- I would not say it is too high but rather high.» (Yes, the level is reasonable)
- Its reasonable. Perhaps more to the lower end of the spectrum, but it all depends on your matlab experience.» (Yes, the level is reasonable)
- I think the level of difficulty is good as it doesn"t require a background in physics.» (Yes, the level is reasonable)
- Especially considering the mathematics.» (No, the level is too high)

14. Did the examination as a whole assess whether you have reached the goals?

34 svarande

No, not at all»		0		0%
Yes, to some extent»		12		35%
Yes, definitely»		22		64%

Genomsnitt: 2.64 (bidrar till totalt genomsnitt/jämförelseindex)

- I gave this feedback before taking the exam.» (Yes, to some extent)
- (This evaluation is done before the written exam) The part about learning when to choose which method has not been part of the home problems. I think maybe the home problems were to strict, that is, there was not really any room for the student to think freely and come up with his/her idea to solve the problem, using either of the methods in the course. I felt that there were more emphasis on how the programs should be written, the coding standard and the pre-written bases for some programs.» (Yes, to some extent)
- Have not read goals, have not taken exam.» (Yes, to some extent)
- I think more emphasize should be put on the actual programming (home problems), and less on the exam. » (Yes, to some extent)
- I think that maybe the limit to get the highest grade is rather high compared to other courses I have taken. I also think that maybe the home problems should yield some more points compared to the exam. It should be taken into account however that I have not written the exam soo maybe I"m not the best one to judge how the points should be divided. (This due to an infected wound :( I will have to try my luck at the next exam)» (Yes, to some extent)
- I think that the home problems should be a lesser part of the grade.» (Yes, to some extent)
- Even the weight of Home Problems could be more than 25p.» (Yes, definitely)
- I felt that the home problems were a more significant part of the course, and believe an adjustment to the scores might be appropriate. Maybe 30p for HP and 20p for exam.» (Yes, definitely)
Genomsnitt totalt för detta stycke: 2.64

Summarizing questions

15. What is your general impression of the course?

Very poor»		0		0%
Poor»		0		0%
Average»		2		5%
Good»		6		17%
Very Good»		26		76%

Genomsnitt: 4.7

- Very good course. I will recommend it.» (Very Good)
- However, I think the workload is too high.» (Very Good)
- Interesting course for sure!» (Very Good)
- Great lectures, and great litterature. Keep up the good work! » (Very Good)
- It has been a very interesting and enjoyable course.» (Very Good)
- A very nice course!» (Very Good)

16. Is there anything you think should be changed until next year?

17. Which areas do you think need more emphasis (or less emphasis) in the course?

- MATLAB Programming»
- Perhaps a bit more on PSO, which seems to perform better than the others (where comparison is possible).»
- More stochastic optimization algorithms could be covered, the course content gets a bit diluted in the second halv of the quater.»
- Always more programming and algorithms :)»
- Slightly less emphasis on classical optimization.»
- My personal opinion is that matlab should be changed to another programming language commonly used in the industry and much faster. Maybe something like c or c++, but it works fine in most of the home problems (except 2.3 and 2.4 which required a lot of computing time).»
- Well distributed overall.»
- I did not have a background in matlab and that gave me a lot of trouble. I would also emphasize more the proofs and the theory behind the algorithms (convergence etc.)»
- Although I appreciate the background information, I don"t see the need for really knowing classical optimisation. Although, if this is seen as a requirement of the course, then maybe more emphasis on this would be good as currently it presumed quite a bit of previous knowledge even though the class had a mixed background and therefore mixed knowledgebase.»
- perhaps less emphasis on classical optimization in the course examination (home problems + exam)»
- More emphasis on ACO.»

18. Additional comments