Chevron Left
Вернуться к Nearest Neighbor Collaborative Filtering

Отзывы учащихся о курсе Nearest Neighbor Collaborative Filtering от партнера Миннесотский университет

4.3
Оценки: 214
Рецензии: 50

О курсе

In this course, you will learn the fundamental techniques for making personalized recommendations through nearest-neighbor techniques. First you will learn user-user collaborative filtering, an algorithm that identifies other people with similar tastes to a target user and combines their ratings to make recommendations for that user. You will explore and implement variations of the user-user algorithm, and will explore the benefits and drawbacks of the general approach. Then you will learn the widely-practiced item-item collaborative filtering algorithm, which identifies global product associations from user ratings, but uses these product associations to provide personalized recommendations based on a user's own product ratings....

Лучшие рецензии

SS

Mar 31, 2019

Thank you so very much to open my eye see more view of recommendation field not only algorithms but use case and many trouble-shooting in worldwide business, moreover interview with noble professor.

NR

Feb 04, 2018

Extremely informative course! It would be great if the assignments are created on python or R in the next season's offering. Thanks for the knowledge!

Фильтр по:

26–50 из 50 отзывов о курсе Nearest Neighbor Collaborative Filtering

автор: Ankur S

Oct 16, 2018

Diverse content that helps in understanding the basic concepts of collaborative filtering. Interviews with people specializing in different nuances of collaborative filteering were very useful.

Some thoughts on what could be improved

Pace of narration. It can be faster

More exercises are needed to get more familiar with the concepts. Each lecture should have a exercise (not just a quiz)

автор: Yury Z

Mar 22, 2018

The topics I am interested in covered by people who definitely has related expertise. But overall quality of the teaching materials expected to be higher. Forum is also a little bit deserted, although contains some critical hints to pass the assignments (such a hints worth to be included in the assignment descriptions itself). I want to support the course, and it is pity to give it only 4 of 5 stars, but I really expect more quality from the course I paid for.

автор: Andrew W

Jan 21, 2018

Thank you for this course -- it opened my eyes to the universal applicability of recommender systems in tech applications.

My feedback is that you could do more to tie the *implementation* to the theory and real-life applications you discuss. You have many great lectures talking about how these systems were implemented, qualitative differences, subtle differences, and interviewing people to give us perspectives. But then the videos on implementation (including working through the equations) are pretty sparse and short. I felt like I'm "on my own" to figure out how to go implement these in real life. The problem sets cover one test case, and that's it. I think you could update the lectures to focus more on different algorithms / equations in different scenarios, rather than just talking qualitatively about them.

Regardless thank you! I deeply appreciate this course and what you've done. I plan to help my Consulting clients directly based on what I learned from you.

автор: Jan Z

Nov 10, 2016

Excellent course providing not only the knowledge of algorithms but also useful insights into developing and maintaining recommender systems. Only thing that could use some work is the assignments. Spreadsheet assignment in week 4 is poorly designed (as evidenced by many forum threads with people not knowing what is it that the authors actually want). Other than that, that was an extremely helpful course.

автор: Edgar M

Oct 25, 2016

Very good content ! Very interesting interviews with expert in the field that shows real examples. However the exercise needs a bit more work to be very useful.

автор: Dan T

Nov 24, 2017

I liked the course, assignment two for item item was so much harder than the user user piece. I really spent all my time fighting excel, rather that working on the problem. it would have been easier to program it in lenskit!

автор: Siwei Y

Nov 27, 2016

Overall , it is a very interesting course.

But I would like to say , that there are too many interviews. I think that it is a little bit difficult for some non-native speaker to understand the main and important things from the interview, because some interviewers talked in a very loose way. So I would suggest our teacher , to summarize the main points of those interview in a better way .

автор: Daniel P

Dec 08, 2017

Rather non-technical, interesting general information, plus voluntary programming assignment which I personally found little bit "bulky". More effort I spent to get familiar with the library than to actually use the collaborative filtering algorithms.

автор: Dino M A

Oct 24, 2016

I think this is very useful for introductory, but it lacks some references for who wants go deeper.

автор: Gui M T

Apr 01, 2019

Much better than the first course, covers more interesting algorithms in more depth. The assignments can be clearer instructions. I also wish the lectures cover actual mathematical examples to work us through the algorithms

автор: Daniil B

Jul 31, 2018

The course itself is interesting, but some of the programming assignments are horribly confusing, what makes you waste your time trying to decipher what the professor really meant. Spreadsheet assignment on Week 3 is the main reason I rate this course so low, and a lot of people on discussion forums agree with me on assignment quality

автор: Kemal C K

Mar 07, 2017

Lessons need more examples.

автор: Ankit A

Jun 21, 2018

Week 4 assignments can do with a bit more clarity.

автор: Anyu S

Apr 29, 2018

Making honours programming exercise in Java is a mistake. Pls consider Python in the future. Assignment for week 4 uses formula differs from the course: wasted many hours that don't benefit learning.

автор: zhenyu z

Feb 21, 2018

the hands-on quiz is not well prepared.

автор: Alberto G

Mar 26, 2018

Assignments are not explained so well on this one

автор: Abou-Haydar E

Aug 28, 2017

The content of the course is actually great, the assignments are a bit challenging which was very interesting. I've learned a lot.

Nevertheless, I didn't enjoy the course much because the support to the course which is inexistent, forum's are almost empty. If you answer a question, you have maybe 1% chance to get an answer from someone, if you open a discussion, it ends up being a monologue...

автор: Daniil

Jun 19, 2019

The course is pretty good, but the spreadsheet assignments are brutal: they are confusing, too tedious and don't have enough information to debug.

автор: Daniel M

Jun 23, 2019

The course material is good, but the course itself is merely okay due to some problems with the assignments that have gone unaddressed for years. The Item-Item filtering assignment solution does not match the formula given in the lectures, and the honors assignments use an outdated version of the code (at one point recommending a package that has been deprecated). Really needs some attention to fix bugs and update the software.

автор: Akash S C

Jul 21, 2019

good introduction to topics and algorithms but very little help provided for the assignment in clarifying doubts in forums and unclear explanations were given for assignments. also not providing option to use any other programming language like python or r to do programming assignment is a big miss. would still recommend this course to get started from basics about reco sys.

автор: Gregory R

Apr 19, 2017

The content of the course is extremely useful, however assignments need review as the exercises results have mistakes and they are not explained very well (missing step by step guidance).

автор: Konstantinos P

Apr 10, 2017

Unfortunately, the content of the course is poor. Too many interviews and some of them are pointless.

автор: Chunyang S

Feb 24, 2017

The content is too basic, and both lectures are too boring.

автор: Jose R

May 27, 2018

Not clear examples in my opinion, and there was same complain made from several user and I never saw a reply and nothing was changed

автор: Alex B

Aug 26, 2019

This course is taught at a really low level. Exercises are in spreadsheets which are more or less useless for practicing scale data applications. Spreadsheets contain information that makes importation into numerical processing software such as Pandas in Python or dplyr in R needlessly difficult and assumes the user can't even apply the distance formula.

Videos contain useful information but require wading through a lot of garbage at a slow pace, not useful for practitioners.

Assignments are poorly worded and some terminology is used questionably or flexibly (see the word "normalization"). Some assignments are so poorly done that there is an ongoing debate on the forums as to whether the autograder is messed up or the assignment instructions are messed up.

The "honors" track programming assignments use some piece of software with questionable generalizability. If I ever see lens kit in my own data work environment I will come back an edit my review but I find it unlikely. Furthermore, Java is not commonly used for data science or machine learning purposes making these assignments inaccessible to many users. Personally, I write in Java but I didn't find it fulfilling to waste my time playing "fill in the blanks" or "guess the library function" which is overall uninstructive.

Quiz assignments show true indications of the poor level of instruction. Recitation of pieces of information buried in 30 minutes videos that can be condensed into 5 are some of the finest examples of bad teaching. Regurgitating information found in required readings shows no level of comprehension of course material and is a severe disservice to students.

I will hope for better general coverage of recommender systems in the future in another course. Ideally using something applicable like Python, Scala (Spark), or even R.