Chevron Left
Matrix Factorization and Advanced Techniques に戻る

ミネソタ大学(University of Minnesota) による Matrix Factorization and Advanced Techniques の受講者のレビューおよびフィードバック



In this course you will learn a variety of matrix factorization and hybrid machine learning techniques for recommender systems. Starting with basic matrix factorization, you will understand both the intuition and the practical details of building recommender systems based on reducing the dimensionality of the user-product preference space. Then you will learn about techniques that combine the strengths of different algorithms into powerful hybrid recommenders....




Really enjoyed the course!\n\nOne suggestion I have is to blend in even more advanced techniques such as using neural networks (e.g. NCF)



great courses! They invite a lot of interviews to let me understand the sea of recommend system!


Matrix Factorization and Advanced Techniques: 1 - 25 / 25 レビュー

by Daniel P


Very interesting topic which I was really stoked to learn but unfortunately this course is missing the deep insight into the algorithms, it explains just one algorithm and its variations. The content is overall to little for 6 weeks course and the honor's assignment has very bad task description with errors and lack of validating possibilities.

by Evaristo C


This is THE BEST course in Recommendation Systems. Probably getting a bit outdated as the field is moving rapidly into deep learning and other techniques, but the problems faced by any recommendation system developer would be the same - ratings measurement, sparse matrices, ranking metrics... And THAT is probably the best contribution of the whole training: not only talking about the methods but pin-pointing the main REAL problems to solve.


by Ankur S


Great course to understand the fundamentals of recommender systems, as well the diversity & challenges of different recommender systems. The interviews with people from relevant academic fields & industry were particularly useful

I really wish that the programming exercises would be in Python. And that more details on how to implement them (the actual modeling of the algorithms etc) was delved into greater detail.

by Su L


It will be great, if we can do honor's track with Python or R

by Bret


The topics that were covered were done well, but I was expecting (and hoping for) way more on probabilistic matrix factorization (I still struggle to understand how it differs from regular MF) and Boltzmann machines. Also, only one assignment in the standard track? And you couldn't auto-grade standard track assignments and quizzes without charging a fee? Weak

by Nicolás A


-some videos need better editing

-should go into more mathematical detail for the matrix factorization techniques

by Daniel M


Content okay, if a little basic. Second honor's track assignment is a complete mess; can't even build properly. Complaints on discussion board have gone unanswered for months.

by Amine D


the way of explaining every concept is very boring

by Keshaw S


Based on my experience with the previous courses in this specialization, I was very positively surprised by the amount and depth of material provided in this course. It covers almost everything that is there to be known. The comprehensive interviews are a big plus point. Also, this course provides guidelines as to how to develop, employ and evaluate a recommender system in real life. I would definitely recommend anyone interested in the field to take this course.

by Blake C


Great course. Professors do an excellent job of breaking down this stuff into digestible bits without losing much substance. Highly recommend. You'll need to do some outside exploration and learning though.

by Hagay L


Really enjoyed the course!

One suggestion I have is to blend in even more advanced techniques such as using neural networks (e.g. NCF)

by Light0617


great courses! They invite a lot of interviews to let me understand the sea of recommend system!

by Joeri K


Great to have people from industry to talk about recommenders as well. Thanks a lot!

by Sanjay K


Awesome course especially for those doing Ph.D in recommender systems

by Ziling C


so very helpful!

by Rahul G


The Hybrid recommenders i.e. Week 4 needs more explanation especially what is Tensor factorization etc, Week 4 was difficult to grasp. Week 5 and Week 6 was informative especially the LinkedIn video and Learning to Rank: Interview with Xavier Amatriain in which a problem was discussed of having that popularity vs Ranking issue.

Thanks it was quite a lot new things to learn ! My Rating 4.5

by Dennis D


Very good. Per closing comments, it probably needs an update (since 2016) as this is active, progressive area.

by Nora P H L


The content is really good, but overall the interviews with experts in the field are the best of this course.

by Alberto G


Programming Assignments are not clear enough and the quiz for the last one seems to be a bit off.

by aankitakaur


Interview with Francesco Ricci

is very knowledgeable about context aware Recommender System.

by LU W


It would be better if explaining how to build latent features

by Siwei Y


内容还是蛮有意思的。 就是 Interview 太多, 而且interview 的人的 上课水平参差不齐, 所以干货 不多。 另外 希望honor assignment 的 参考结果 应该比较 reproducible, 否则很难知道自己的 code 错在哪里。

Content is really interesting, But there are too much interviews, to give students more systematical impression. Besides that , I hope that the example output of honor assigment should be more reproducible , otherwise it is hardly to know if our code is corrrect.

by Kemal C K


Few quizes. Easy tests. We need more information on the core techniques.

by Moustafa M


The HWs for the Honor track had mistakes



Could have been better