Aktuella utvärderingar

Visa resultat

Här kan se resultatet från utvärderingen och exportera statistiken till ett annat program. Det går också att göra en enkel filtrering genom att klicka på svarsalternativen och kommentarerna eller en avancerad filtrering genom att använda knappen längst ned.

Algorithms for machine learning and inference, lp3 VT11, TDA231/DIT380

Status: Avslutad
Öppen för svar: 2011-03-18 - 2011-03-24
Antal svar: 29
Procent av deltagarna som svarat: 37%
Kontaktperson: Rebecca S»
Utbildningsprogram som genomför enkäten: Informationsteknik 300 hp

Your own effort

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

29 svarande

At most 15 hours/week»7 24%
Around 20 hours/week»16 55%
Around 25 hours/week»3 10%
Around 30 hours/week»2 6%
At least 35 hours/week»1 3%

Genomsnitt: 2.1 (bidrar till totalt genomsnitt/jämförelseindex)

2. How large part of the teaching offered did you attend?*

29 svarande

0%»1 3%
25%»1 3%
50%»4 13%
75%»8 27%
100%»15 51%

Genomsnitt: 4.2 (bidrar till totalt genomsnitt/jämförelseindex)

Genomsnitt totalt för detta stycke: 3.15

Goals and goal fulfilment

3. How understandable are the course goals?

29 svarande

I have not seen/read the goals»6 20%
The goals are difficult to understand»1 3%
The goals give some guidance, but could be clearer»9 31%
The goals clearly describe what I am supposed to learn»13 44%

Genomsnitt: 3

4. Are the goals reasonable considering your background and the number of credits?

25 svarande

No, the goals are set too low»0 0%
Yes, the goals seem reasonable»24 96%
No, the goals are set too high»1 4%

Genomsnitt: 2.04

5. Did the examination assess whether you have reached the goals?

29 svarande

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

No, not at all»0 0%
To some extent»4 13%
Yes, definitely»22 75%
I don"t know/have not been examined yet»2 6%

- Some of the contents mentioned on the studentportal wasn"t included in the course.» (Yes, definitely)
- It"s hard to know since the grading criteria are very vague.» (I don"t know/have not been examined yet)
- The exam was good and tested almost everything, but there should be more exercises in advance allowing for practice on the type of questions that appeared on the exams, apart from the home exercises.» (Yes, definitely)
- The course is to theoretical» (To some extent)

Teaching and course administration

6. To what extent has the teaching been of help for your learning?

29 svarande

Small extent»3 10%
To some extent»4 13%
Large extent»11 37%
Great extent»11 37%

Genomsnitt: 3.03

- Since the lecture notes wasn"t available before the lecture it was very hard to prepare for the lectures.» (Small extent)
- In my opinion, more practical examples and applications should be cover for the different fields, especially for the more abstract concepts.» (Large extent)
- I think that the explanations are too mathematical. Mathematical reasoning is great but it should never replace the intuitive explanation which, at least for me, is much easier to understand. Examples helped a lot as well.» (Large extent)
- Great teaching!» (Great extent)
- Peter Damaschke is a great lecturer! Things that I have seen in other courses were explained here in a much more easy way. However, he should be careful when using words like "obviously" and "trivialy". Those can feel like a slap in the face when you are struggling to keep up.» (Great extent)
- Awesome lectures.» (Great extent)

7. To what extent has the course literature and other material been of help for your learning?

29 svarande

Small extent»1 3%
Some extent»11 37%
Large extent»9 31%
Great extent»8 27%

Genomsnitt: 2.82

- Could be better more resourses, and different ways of presenting the topics, since the very theorical view presented is very important, but at the begining it makes harder to start to grasp the concepts» (Some extent)
- The lecture notes summarized the course lecture in a good way. The exercises were most helpful.» (Some extent)
- The course notes were really great! The Mitchel"s book was a little complicate.» (Some extent)
- I only used the lecture notes, not the book.» (Large extent)
- The lecture notes is very good for learning.» (Large extent)
- Lecture notes were great but sometimes very compact. Examples and "real calculations" with numbers would have been a great help.» (Large extent)
- Although a more updated book would be appreciated, not that Mitchell"s book is bad though.» (Great extent)
- Since the lectures didn"t do the trick I had to rely almost entirely on the book.» (Great extent)
- The lecture notes are priceless.» (Great extent)

8. How well did the course administration, web page, handouts etc work?

29 svarande

Very badly»1 3%
Rather badly»0 0%
Rather well»8 27%
Very well»20 68%

Genomsnitt: 3.62

- The website is not good at all. You cannot update entries in a list as "news". It"s almost impossible to see if there is something new. You also need to use links at the top of the page. It is very hard to get an overview of the site using only horizontal rulers and small header. There are professors at Chalmers who specify in user interfaces, you should contact one of them, Olof Torgersson (oloft@chalmers.se) for example.» (Very badly)
- Changing the hand in time for the exam at last minute is pretty bad, but at least it was an extension.» (Rather well)

Study climate

9. How were the opportunities for asking questions and getting help?

29 svarande

Very poor»1 3%
Rather poor»1 3%
Rather good»5 17%
Very good»17 58%
I did not seek help»5 17%

Genomsnitt: 3.82

- No scheduled lab times...» (Very poor)
- I will come to that later» (Rather poor)
- Great response time for both home problems and exam questions.» (Very good)
- Even though I"ve never used it, it feels comforting when a teacher says that you can book consultation times with them.» (Very good)

10. How well has cooperation between you and your fellow students worked?

29 svarande

Very poorly»0 0%
Rather poorly»5 17%
Rather well»5 17%
Very well»12 41%
I did not seek coopeation»7 24%

Genomsnitt: 3.72

- Not much interaction encouraged by the structure of the course» (Rather poorly)

Summarizing questions

11. What is the hard part of understanding the notion of inductive bias (if you found it hard), and how could that be supported?

- Its not hard»
- I didn"t find it hard»
- Examples and repetition - the mother of all (human) learning.»
- It can be supported by asking more questions about it, in the hand-in exercises.»
- When you start the couse it feels like it contradicts the learning. Without major help, the machine cannot learn anything. Maybe this can be helped by showing some of the inductive biases used later in the course and their areas of use. Maybe then can students understand that a bias is not "giving the algorithm the answer" but mereley limiting the hypothesis space.»
- The difficult part was learning the difference between restricting the hypothesis space and just removing hypothesis or removing parameters. (learning that abstraction) I think that it could be supported by adding more toy-examples like the one of the lamp. Or/and maybe counterexamples of what an inductive bias is not.»
- In my opnion, the weka tutorials (not the assignment we had to hand in) and the support in the practical were crutial to my understanding of inductive bias.»
- That EVERYTHING, no matter how obvious or inconsequential it may seem, is part of the inductive bias. Just stress this a bit more and make it more explicit by pointing out all the inductive bias in a really trivial example?»
- I think the introduction was too focused on the concept learning example. That was a bit misleading as a lot of people have only understood inductive bias in that context.»
- No changes necessary.»
- I think I got it quite early.»

12. What is your general impression of the course?

29 svarande

Poor»0 0%
Fair»1 3%
Adequate»4 13%
Good»16 55%
Excellent»8 27%

Genomsnitt: 4.06

- Although I appreciate the theoretic foundation of the course, this is an introductory course and I would expect more algorithms to have been mentioned (nothing from genetic algorithms, analytic learning was mentioned although its on the studentportal)» (Adequate)

13. What should definitely be preserved to next year?

- The weekly exercises, it forces the student to work throughout the whole course. But perhaps one could extend it to have a practical task in each. I.e, implements these algorithms (many aren"t all that hard) and evaluate this, comment on this, verify the following theoretical relations etc.»
- Home problems, they create a good incentive for studying outside of the lectures.»
- the exercises»
- The way the exercises are managed, since it is possible to inprove your knowledge, and actually they let the student identify gaps in the concepts»
- The exercises.»
- The teacher!»
- Peter Damaschke.»
- Hand-ins and take-home exam»
- The lecture notes. The practical session.»
- The excercise sets, and Peter"s excellent lectures.»
- The exercises!»
- The lecturer and lecture notes.»
- Both Home assignments and WEKA lab.»

14. What should definitely be changed to next year?

- It would be good if the grading criteria was a bit clearer or maybe complemented with a grade that is received when an exercise has been passed. I have now completed my exercises and the exam but I have no idea whether I meet the criteria for 4 or 5, i believe a meet the criteria for 3 but I am not sure. »
- Re-include GA, Analytical learning and perhaps other algorithms»
- The lab was hard to grasp (bad labb-pm) and not that giving.»
- More support resources for the lecture, motivate the concepts in a more familiar way before going deep in the formal definitions»
- Preferably introduce home study exercises with answers to improve overall understanding.»
- Lab deadlines. First deadline could be set earlier... you should at least provide a longer time between first and final deadline, in my opinion.»
- course. The worst part is that the things he claims is "obvious" is rarely obvious at all. Take this sentence from lecture 5: "One obvious implementation of the ensemble idea for decision trees is known as Random Forests". Sure it is easy to understand once it has been mentioned, but "obvious" means that everyone should be able to see it without it being mentioned. Please Peter, stop.»
- The weka lab can be modified to be less of a "press buttons and observe results" lab. »
- I could be good to have some implementation too.»
- The Weka excercise felt a little out of place, perhaps do something to spice it up (for example, a more open task like "find correlations between these search patterns, then motivate why you chose the method you used?")»
- Home exam.»
- The teacher needs to stop saying "obvious", "trivially" and "clearly". In just the written parts of the course (Exam solutions and lectures) the teacher says these words a staggering 46 (!!!!) times. Forty six, I want to say it again because I just can"t believe it. A teacher saying that something is "obvious" is saying to the student that if he doesn"t understand it then he"s stupid. A teacher should never say that a student is stupid. I"ve felt that the teacher has called me stupid several times during the course. The worst part is that the things he claims is "obvious" is rarely obvious at all. Take this sentence from lecture 5: "One obvious implementation of the ensemble idea for decision trees is known as Random Forests". Sure it is easy to understand once it has been mentioned, but "obvious" means that everyone should be able to see it without it being mentioned. Please, stop.» (den här kommentaren har blivit redigerad i efterhand)

15. Additional comments

- Great course!»

Additional comments

Genomsnitt totalt för alla frågor: 3.15
Beräknat jämförelseindex: 0.53

* obligatoriska frågor

Kursutvärderingssystem från