Visa resultat

Algorithms, period 1, 2007

1. Your study programme

2. Your workload in the course:

low, I would even like more»		0		0%
low»		0		0%
just fine»		9		37%
high, still OK»		13		54%
too high»		2		8%

Genomsnitt: 3.7

3. How would you assess your previous background in algorithms and data structures?

strong»		3		12%
suitable»		16		66%
weak»		5		20%
none»		0		0%

Genomsnitt: 2.08

4. Course goals are:

modest»		1		4%
just appropriate»		7		29%
challenging, but OK»		12		50%
too ambitious»		4		16%

Genomsnitt: 2.79

5. Did the course reach the declared goals?

more than that»		4		16%
yes, fully»		12		50%
to a large extent»		7		29%
partially»		1		4%
barely»		0		0%

Genomsnitt: 2.2

6. Will it be useful what you have learned in the course?

yes, much»		17		70%
somewhat»		5		20%
hard to predict»		0		0%
I do not think so»		2		8%

Genomsnitt: 1.45

7. Did you get interested in learning more about algorithms and their applications?

yes, much»		11		45%
somewhat»		8		33%
not more than before»		4		16%
no»		1		4%

Genomsnitt: 1.79

8. About proving properties of algorithms (correctness, complexity), which statement is closest to your opinion?

It is obviously important to do it.»		14		58%
It is important, but I hardly have to do it by myself»		7		29%
Why bother. Implement and test, then you will see.»		3		12%

Genomsnitt: 1.54

- i find it very hard to start proving things. to know where to start. » (It is obviously important to do it.)
- I have taken the course earlier but didn"t pass the exam.» (It is obviously important to do it.)
- Why bother, unlikely that we will EVER use this information. The algorithm course should be changed to some advanced data structure course instead.» (Why bother. Implement and test, then you will see.)

9. How many lectures did you attend?

13-14»		16		66%
9-12»		6		25%
5-8»		2		8%
1-4»		0		0%
0»		0		0%

Genomsnitt: 1.41

10. Did you think about the problems before classes, as recommended?

always»		2		8%
often»		6		25%
sometimes»		9		37%
never»		7		29%

Genomsnitt: 2.87

11. How much did you understand already in class?

(almost) everything»		1		4%
most of it»		10		41%
part of it»		6		25%
a bit, I got an idea»		5		20%
(almost) nothing»		2		8%

Genomsnitt: 2.87

- I have taken the course before so this was a recap for me.» ((almost) everything)
- There was never any incentive doing problems outside class since no key was given to selected excercises. This was really a problem. There was no point in doing any work because you couln"t get confirmation that you were on the right track.» (most of it)
- Very good lectures!» (most of it)
- Having lecture notes so that i don"t have to write and draw would help very much. This way i could focus on the problems.» (a bit, I got an idea)
- Have you tried understanding his English??? it is impossible, now put some algorithm on it and do the math!!!» ((almost) nothing)

12. Did you buy the course book?

yes»		17		70%
no, too expensive»		7		29%
no, other reason»		0		0%

Genomsnitt: 1.29

13. Course material in general

very good»		5		20%
good»		11		45%
OK»		6		25%
bad»		0		0%
very bad»		2		8%

Genomsnitt: 2.29

- Book is sufficient for self study.» (very good)
- the book was exceptionally good, but it helped only to understand the algorithms and it gave means to analyse them and prove (in)correctness. but id did not do much to help solving problems. eg. how do I figure out what is the merging step in some divide and conquer algorithm.» (very good)
- I really liked both the book and the lecture notes.» (very good)
- Good lecture notes, though could use some figures / pictures» (good)
- I like that you posted the problems before the lecture, but it was a bit unclear which problems you were going to address. Right now there are three files containing problems, but we have many more lectures. Perhaps split them up so we know exactly what material you will be presenting. One link for each lecture would be nice. The lecture notes you provide are nice. But most of these algorithmic problems become much clearer with an image or two to explain things. Mathematical expressions tend to be a bit abstract for the average computer science student. If you would include some of the drawings from the lectures into the lecture notes you would have a very competent book on your hands. Play your cards right and you just might some money out of this.» (good)
- Lecture notes were well written but too "text-dense". Some sketches would be a good addition.» (good)
- very good lecture notes. the book was (as always) expencive and not very necesary (the only time i used it was at the exam to review dijkstras algorithm). hopefully i will have more use for it in the advanced course» (good)
- Poor overview and structure in the book» (OK)
- Answers, on book exercises would be helpful.» (OK)
- Lecture notes are very hard to read, and even harder to understand. Use more pictures/diagram/models,... not only TEST to explain an algorithm.» (very bad)
- The lecture notes are impossible to understand. » (very bad)

14. Which subjects are particularly difficult?

(på denna fråga var det möjligt att välja flera svarsalternativ)

basic notions»		2		8%
heuristics»		1		4%
greedy»		1		4%
dynamic programming»		6		25%
divide-and-conquer»		15		62%
graph algorithms»		4		16%
reductions»		14		58%
P and NP»		10		41%

- This is a mathematical course on the advanced level. They are hard by definition. If you put your mind to it you could make any of the above "subjects" into a really hard problem.» (dynamic programming, divide-and-conquer)
- Non-obvious divide-and-conquer algorithms are difficult to grasp. "Why divide in 5 and not in 3" and similar questions are not simple. Also, the recurrence relations associated with D+C can be challenging. Thank God for the Master"s theorem.» (divide-and-conquer, graph algorithms)
- divide and conquer - as the technique is so simple then the problems are made complicated through other means which does not have anything to do with how well I understand the principle of divide and conquer. » (divide-and-conquer)
- It is hard to see what problems are similar if you dont know it forehand.» (reductions)

15. Which aspects are particularly difficult?

(på denna fråga var det möjligt att välja flera svarsalternativ)

motivations, applications»		3		12%
ideas, algorithm design process»		11		45%
correctness proofs»		22		91%
data structure details»		2		8%
complexity analysis»		7		29%
apply high-level descriptions to concrete instances»		6		25%

- I dont know how much explanation is enough to be matemathically correct.» (correctness proofs)
- most of the algorithms were more presented than designed in the classroom. exception being greedy algorithms, and this process seemed like trial and error.» (ideas, algorithm design process)
- The hard thing for me is to come up with ideas to solve problems. Sadly I have not practiced this very much.» (ideas, algorithm design process, correctness proofs)

16. Effect of programming assignments:

I got a better understanding of the algorithms»		15		62%
I got at least some new insights»		8		33%
I did not learn anything new»		1		4%

Genomsnitt: 1.41

17. Did you get the necessary help?

I did not seek help.»		12		50%
yes, OK»		6		25%
yes, but could be improved»		4		16%
insufficient»		2		8%

Genomsnitt: 1.83

- more programming assignments, maybe with combinations of the methods.» (I did not seek help.)
- Already done the programming assignments earlier.» (I did not seek help.)
- If you asked Libertad she would sometimes refuse to help you. Sure, we are suppose to think for our selves. But if you have spend a week on a problem and then come to her to discuss it and you leave just as confused as before simply because she doesn"t feel like it, well, it sort of makes you a bit angry.» (yes, but could be improved)
- two teachers, for 100 student, it is a joke.» (yes, but could be improved)
- Only one exercise session each week is just stupid for this kind of course. at least 2 or even more sessions are required.» (insufficient)

18. How many exercise sessions did you attend?

5-6»		11		45%
3-4»		6		25%
1-2»		5		20%
0»		2		8%

Genomsnitt: 1.91

19. Quality of exercise sessions (were they helpful and instructive)?

very good»		2		9%
good»		7		31%
OK»		8		36%
bad»		4		18%
very bad»		1		4%

Genomsnitt: 2.77

- Didn"t have time.» (?)
- depended on teacher.» (good)
- This is where the actual learning takes place.» (good)
- The one with Peter was the by far best one.» (OK)
- The one with Peter was helpful. I like it more if there is no time for trying to solve exercises alone at first. One could do it at home. Time for exercise session was really bad, we had lectures from 8am and the ex session was the last one, ending at 17.» (OK)
- Well, I attended both Libertads and Haralds sessions. Sometimes it just made things worse and sometimes it actually helped. The "layout" of the sessions appears to be a bit undefined, which sometimes resulted in a bad overall experience.» (OK)
- Answers, on book exercises would be helpful so one can work through the exercises that there not done during the exercise session.» (OK)
- What is the meaning of just waiting for one hour before they help us solve the algorithms? The thinking part we can do at home, just solve them so can learn something.» (bad)
- If the solutions for the exercises where posted on website it would have been as helpful. there was very little said how to approach those problems.» (bad)
- There wasn"t time to get answers to all questions, and the we didn"t get the answers for att of the suggested exercises.» (bad)
- The teacher spend lots of time doing nothing.» (very bad)

20. If there were more exercise sessions, would you actually attend more?

yes»		16		69%
no»		7		30%

Genomsnitt: 1.3

21. Would you suggest a different form of problem-solving training?

no, OK as it is»		15		65%
yes (comment)»		8		34%

Genomsnitt: 1.34

- What is the meaning of just waiting for one hour before they help us solve the algorithms? The thinking part we can do at home, just solve them so can learn something.» (yes (comment))
- I would suggest more problems to solve and hand in. It"s like in programming, you cant be and armchair programmer it the same way you cant be an book reading algorithms wizard. You have to use these techniques to fully understand them and their variations. How about one assignment each week? And please give constructive help to the ones who seek it, otherwise you end up with upset students. Also it would be very nice with solutions to some of the exercises in the book. Right now you have to attend all the exercises to get a few answers, but it doesn"t really allow you to study on you own in a more constructive way. I looked on the internet and could not find any solutions to the text book problems. I can understand if you dont have time do make an full set of solutions to all the problems in the book. But perhaps you could pick a few that are good and give us something to work with that doesn"t require us to ask you directly?» (yes (comment))
- I would not mind more open programming assignments. They were very simple since the full details of the algorithms already were given.» (yes (comment))
- Make the exercises two/three times each week.» (yes (comment))
- Don"t keep the students away from the key. That may be a good idea in elementary school, but this is a university.» (yes (comment))
- I was very strong algorithms course but not so much problem-solving. And I think the exam did not test very well whether I acquired the material. eg. fair boss problem. If id did not think that, if I fill half the time as well as possible then I get correct solution - to figure out that had nothing to do how well I knew the dynamic programming (to make it worse next two questions depended on figuring out that).» (yes (comment))
- More exercises from the book that has answers so that you can do exercises at home or in groups. There is no meening in doing exercises if you won"t get the answers so that holds back the chance of understanding and learning the parts of the course.» (yes (comment))
- I dont know since i didn"t attend any excercise.» (yes (comment))