Chevron Left
Вернуться к Машинное обучение

Отзывы учащихся о курсе Машинное обучение от партнера Стэнфордский университет

Оценки: 151,186
Рецензии: 38,564

О курсе

Machine learning is the science of getting computers to act without being explicitly programmed. In the past decade, machine learning has given us self-driving cars, practical speech recognition, effective web search, and a vastly improved understanding of the human genome. Machine learning is so pervasive today that you probably use it dozens of times a day without knowing it. Many researchers also think it is the best way to make progress towards human-level AI. In this class, you will learn about the most effective machine learning techniques, and gain practice implementing them and getting them to work for yourself. More importantly, you'll learn about not only the theoretical underpinnings of learning, but also gain the practical know-how needed to quickly and powerfully apply these techniques to new problems. Finally, you'll learn about some of Silicon Valley's best practices in innovation as it pertains to machine learning and AI. This course provides a broad introduction to machine learning, datamining, and statistical pattern recognition. Topics include: (i) Supervised learning (parametric/non-parametric algorithms, support vector machines, kernels, neural networks). (ii) Unsupervised learning (clustering, dimensionality reduction, recommender systems, deep learning). (iii) Best practices in machine learning (bias/variance theory; innovation process in machine learning and AI). The course will also draw from numerous case studies and applications, so that you'll also learn how to apply learning algorithms to building smart robots (perception, control), text understanding (web search, anti-spam), computer vision, medical informatics, audio, database mining, and other areas....

Лучшие рецензии

18 июля 2019 г.

Amazing course. It gets deep into the content and now I feel I know at least the basics of Machine Learning. This is definitely going to help me on my job! Thanks Andrew and the mentors of the course!

26 мар. 2018 г.

Very well structured and delivered course. Progressive introduction of concepts and intuitive description by Andrew really give a sense of understanding even for the more complex area of the training.

Фильтр по:

426–450 из 10,000 отзывов о курсе Машинное обучение

автор: Tan M

30 окт. 2016 г.

It has been a great learning experience taking this Course. I am currently taking an advance version of Machine Learning in my school, and this course on Cousera has definitely provided me the basic and essential knowledge in tackling more advance machine learning problems in school.

To the mentors, thank you for answering my questions that i have posted in the forum. Just a little feedback, i hope that there will be different mentors tackling on different weeks' problems (Spreading the workload ..maybe). In this way, answers to some of the questions can be more detailed.

And i also hope that some days the errata in videos can be corrected even though these are minor that students do not need to refer so much to the errata page while watching the video lectures.

Overall, the course is great and i will definitely recommend this to my friends! I hope that one day Coursera will have an advanced version of the class! (etc. Machine Learning II)

автор: Brian T S

13 февр. 2018 г.

Professor Ng provides an extremely accessible overview of AI techniques. The math does seem a bit imposing, but anyone with a background in pre-calculus or higher should be able to "get it" if they sit down and work it out. I took a similar course at the University of Texas in the 90s and this was presented in a much more understandable manner. It might have been beneficial to work on a complete AI programming assignment at some point, or at least accomplish more of the coding, as most of the assignments required completion of the more trivial aspects of the technique. Also, I was a little surprised that search (tree traversal) was not addressed. Maybe that's too old hat, or covered in a graph course. There were some frustrations with Coursera not working correctly (unable to submit assignments due to broken URL forwarding, broken Latex rendering which is still not 100% working), but I really like the site when it works. High marks for Professor Ng!.

автор: Tom M

30 янв. 2018 г.

I reviewed many courses before taking this one and I'm sure that I made the right choice. This course covers the underlying mathematics of how the various learning algorithms work. Understanding at that level is essential to designing and debugging machine learning systems rather than just applying rote techniques or blindly calling library functions in an ML framework.

I found parts of the course challenging as I'm not a great mathematician but I'm very glad I persevered. The pace and structure of the course were just right. I've just got one tiny gripe. I suffer from multiple hearing problems so at times I needed to turn on the captioning. The problem with automatically generated captions is that they struggle (i.e. get wrong) precisely the same words that I am struggling with, so for me at least, the captions were useless.

Overall, I would describe this as the defininitive 'must-do' course for anyone looking to get involved with ML applications.

автор: Joseph M

7 февр. 2016 г.

Fantastic. Andrew Ng is a naturally charismatic teacher with a knack for anticipating issues which his students may encounter and assuaging them before they become sticking points for later understanding. By their nature, online courses cannot benefit from students asking questions of their instructors so it is doubly important that instructors be aware of areas which may confuse students and take anticipatory action to avoid this- this is only one of Ng's strengths. Beyond this, Ng is simply an enthusiastic instructor whose passion for his subject is contagious. He also conveys a genuine sense of understanding the student's process of coming to grips with more difficult portions, often explaining what has confused him before (though, given his expertise, one may wonder just how much these areas actually give him difficulty). All things considered, the biggest disappointment is that there are not more courses available with Ng as the instructor.

автор: Leonid B

10 мая 2020 г.

First of all - thank you very very much to Andrew Ng !!! There are some things which can be improved - as always but in general, I think it is perfect. I am a hight Energy Physicist and I have some reasonable good knowledge of mathematics. Some of the explanations are so good and so clear that I would use them while explaining quartum mechanics ))))))) and this is not a joke. I was inspired by Andrew Ng and had read partially (not entire) the Ph.D. thesis of Andrew Ng. Thank you one more time for all the algorithms you have explained here. Thanks to you I have: 1) my own spam classifier with my own dictionary - for this, I use bash and c++ library, 2) I have my own list of movies and I make these lists for friends :-) !!!!! 3) and finally, I have the vector components of my face and faces of my friends in the space of celebrities )))))). I wish you and your closes surrounding to keep healthy happy and focused on the things you like to do !!!!

автор: Tony W

25 нояб. 2016 г.

Excellent course! There are many small mistakes throughout, but these are addressed in the errata provided. In addition, a few mentors were constantly providing helpful feedback, even to questions covered elsewhere (such as the errata). I came to the course with a strong statistics background, so one o the very helpful things for me was learning the machine learning terminology for things I was already familiar with from statistics. One final point - the course is structured so that you are improving your general programming skills as you proceed. You begin by doing the simple for-loop implementation but will later see/do the more efficient vectorization. This is incredibly helpful for those who do not come with a solid basis in linear algebra / matrix algebra, since you do the intuitive/"easy" version first but later develop more efficient coding which you now understand because you did the intuitive/"easy" version earlier in the course.

автор: Matthew E H

1 июля 2020 г.

Andrew Ng is an outstanding instructor. This course takes a breadth-first approach, first glossing over the details to describe how various ML techniques are used, then coming back to describe how the underlying algorithms work. This approach worked very well for me. The intermixing of video lectures and straightforward programming exercises worked well to cement the concepts. I have a much better understanding of the ML domain than before I started. Over the past few years I have read a number of books attempting to come up to speed in the fundamentals of machine learning, but I found each lacking, either going too deep into the underlying mathematics before describing how various techniques could be used, or providing only surface level descriptions with no real world examples. I am so happy that I finally put the time into this class, I would recommend it to anyone wanting to come up to speed in the fundamentals of machine learning.

автор: Anant B

11 янв. 2017 г.

This course is very well organized and exposes students to some fundamentals of Machine Learning and practical applications of them. Assignments and guidance to complete them from resources, tutorials and test cases are absolutely helpful to learn these basics and their application with an hands on approach and gives students much more confidence on what he is learning and has been able to absorb. There is a lot of complex materials covered in this class which at the beginning looked fairly insurmountable but working the Assignments with help from resources in form of tutorials, test cases to run and of course the very valuable forum discussions and moderator Tom M's continuous help made learning a lot easier than I thought. Finally thanks to Prof Andrew Ng for offering such an well organized course like Machine Learning through video lectures, lecture notes (pdf) and exercise files to make learning more meaningful and much easier to absorb.

автор: Pedro G

14 авг. 2020 г.

I always had the interest in learning about ML topics in general, because of the unlimited capabilities of powerful algorithms that make possible the processing of data that, for us humans, results generally in something non-comprehensible, but it brings out marvelous results from scratch! This course offered me not only the opportunity to increase my abilities with matrices development in OCTAVE/MATLAB (and many, but more many ways in that it is possible to do calculations with a focus oriented to a tremendous performance), but in essence and most relevant, the right path to learn something of so much utility and admiration (and to be honest, very complicated at beginning to get familiar with) in such a high-skilled professional manner for explaining all involved themes. Every scenario must have a context in order to be understood, and this course accomplish with it at utmost level. I assure you will be very satisfied with these lessons :)

автор: Simon C T

16 июля 2020 г.

Firstly, I will like to say THANK YOU! to Mr. Andrew Ng. For putting in so much work and effort into such an amazing online program that provided me on a great introduced into Machine Learning.

I am passionate on learning and understanding first principles is something that I value. Knowing how certain concepts are formulated and developed helps me to visualize how things work. During this course Mr. Andrew Ng made complex technique very straightforward and comprehensible. This comes not only from having very good knowledge and experience but most importantly the ability to communicate to his students. Thank You Andrew Ng. for conducting this EXCELLENT course. I intend to continue to use and review the material and videos from this course as I develop my knowledge in Machine Learning. Truly loved the videos, especially the very last one and not because it didn't have a quiz in it .. :)

Simon Chan Tack (Machine Learning Engineer & student).

автор: Bruce M

1 янв. 2017 г.

Great content and explanations. Exercises guided hands-on efforts through all of the major algorithms and concepts, including very useful tools such as training rate analysis, cost and gradient descent verification, etc. Exercises were well structured to focus on key concepts - i.e. didn't have to spend a lot of time sweating the details of loading datasets, plotting results, ... The code I take away is an excellent foundation with which to explore other datasets with confidence.

I was originally a bit dubious about Octave-based exercises based on some reviews I'd read from others (who were more R or python-centric), but I found it to be an excellent choice given the linear algebra underpinnings of all of the algorithms and exercises.

The linear algebra foundation and vectorization implementations I think better prepare me in approaching large scale ML problems leveraging existing / emerging linear algebra and ML libraries.

Thanks again!

автор: Bill F

28 мая 2019 г.

Before taking this course, I took the Coursera Applied Data Science specialization which included a higher level view of ML and applications to NLP and Social Networks. That series did not delve into ML algorithms but focused on the application of ML libraries and understanding the some of the different learning classifiers and regressors. This course is an excellent complement to that specialization, the former gave a broader view of the landscape, and having that training, it gave me much greater understanding of and appreciation for what these algorithms were doing. It stretched my non-software background but I was able to learn enough to effectively complete the programming assignments. I know the whole machine and deep learning field is complex and challenging, and for a novice like me this is not going to turn me into an expert (as Prof Ng suggested) overnight, but I feel like this gave me a solid footing to further pursue it.

автор: Justin C

20 февр. 2019 г.

There is a reason this course is so well regarded!

Andrew Ngs teaching method is to take a complicated concept and break it down into steps. At the end of each sections videos he presents the entire concept, at which point you intuitively understand the pieces.

When first reviewing this course I thought he didn't explain things very well, but I was wrong. The truth is, he's taken very complicated concepts and made them as simple as they could possibly be.

I would recommend this course to anyone that wants a practical understanding of Machine Learning.

I personally studied math in preparation for this course and it was helpful. It is possible to complete this course with almost no pre-requisite knowledge like math or basic coding, but you will have to do some research on your own time to fill in the gaps. I would highly recommend completing the assignments to gain a better understanding of the concepts.

You will be happy you did this course!

автор: Nicola G

20 июня 2018 г.

Very good introduction to machine learning, Andrew is a great teacher, always clear and slow enough to be able to follow everything in detail (sometimes even too slow). I like the little questions in the middle of the explanations and the quiz, well thought and almost always touching very important concepts that could be overlooked. Programming exercises are well designed, I never had problems understanding the assignment or submitting my code. Just one side note: most of the code is already implemented and sometimes there is very little the student has to do (literally 2/3 lines of code). It would be great if students had more "freedom to fail" and figure out how to fix their code. However, I realize this would be less appealing for most of the students.After all, great course, I would definitely recommend it to anybody interested in the topic without previous knowledge! I am going to take the deep learning specialization courses now!

автор: Rostislav D

6 июля 2018 г.

An incredible introductory Machine Learning course! Everything you need to get started with developing Machine Learning systems for practical applications. Programming exercises provide a solid framework for creating your own ML algorithms. Videos are to the point and give all the necessary mathematical background without going too deep into the theory behind the mathematics, but just enough to create the working implementation of the algorithm at hand. Quizzes during the videos and at the end of the week help to solidify the newly learned concepts and techniques. The lecturer, Andrew Ng, deserves a medal not only for creating this course, which has probably been a lot of work to begin with, but also for clear explanations and overall, for being a great mentor throughout this wonderful journey. What can I say, I can only wish to take more courses by Andrew to learn more about the subject and gain even more expertise on the subject!

автор: Krishna P

26 дек. 2017 г.

Till I jointed this course, I thought ML was one of the toughest to learn and not the for the guys like me, who is out of Math for a decade. But the way Andrew Ng, takes this course forward, it feels like easier than learning how to code for finding prime nos. Yes, I am a big fan of Prof. Andrew Ng, and already consider as my ML guru. So much motivation I got from this course that, I have already started taking his Neural Network and DeepLearning specialization course from coursera. He mastered the art of teaching tough concepts in easy to understand methodologies. With 95% passing grade, I can confidently recommend this course to anyone.

One word of caution. Don't join this course, because someone recommended and later complain that its not a good course, if you cannot put your honest hard work and dedication. Don't join the course out of curiosity, join with passion and I bet you that you see the end of 11th week of this course.

автор: Sayan G

12 окт. 2018 г.

Firstly, I would like to thank Coursera and Professor Ng for making this course available for people like myself, who have been observing recent shifts in the information technology industry from within. It has allowed me get to know, albeit at a very basic level, the machine learning algorithms being used currently in the industry, and also take beginner steps towards being able to implement them using a programming language.

Secondly, the course acts as a springboard for more specialized programmes which one might want to train oneself in. Personally, I would like to know more about Neural Networks on the Supervised Learning side, and Recommender Systems on the Unsupervised Learning side, and the information already presented herein allows me to get a head start.

I intend to make extensive use of the information I have gathered from this course, as I work towards my goal of becoming a Data Science professional.

Best Regards,


автор: Li Z

26 июля 2017 г.

Highly recommended to anyone who wants to understand what machine learning is about. This is by far the best teaching material available online that I know as an introductory class to machine learning. I know some python programming and very little C before taking it; I tried to read the codes on Kaggle website to understand their projects, but only found myself not understanding anything when it comes to data analysis with machine learning. My friend recommended this class to me and I am glad I spent three months to study it. Never done any matrices calculations before, but it is not hard to understand it; and I have forgot most of my college math (major in basic science research in the past), with some help (online and friends' ) again that's not difficult to understand the content either. Now I am excited to learn some more advanced machine learning skills and hope to do some projects for practice. Thanks for this great course!

автор: Dimitar D

6 дек. 2015 г.

The course provides a sizeable amount of pretty cohesive material, which can still be understood by non-CS students. It's very practical and it includes a very nice mix of quiz tests and great MATLAB/Octave programming assignments. After going through the assignments I started wondering about other problems which data sets I can plug with small modifications into the completed solutions. Andrew Ng keeps a great balance between explaining important details and skipping over parts that require straying too much from the main topic of the lecture. I still don't have very deep or broad knowledge in the Machine Learning domain, but it feels like the course doesn't miss anything of the fundamentals.

Overall, I'd definitely recommend the course to CS students, high-school students with interests in the computer science area and even specialists in other areas with some knowledge in linear algebra with interest in the AI and ML domains.

автор: Guo X W

6 июля 2020 г.

Had high expectations prior to taking this course because of the rave reviews. This course truly exceeded expectations. Andrew Ng explains ML concepts in a simple, intuitive manner. He tells you exactly what you need to know, and yet he provides optional videos for students who are more mathematically inclined. The programming assignments are very well-designed, and serve to reinforce understanding of key topics. I also appreciated the nice balance between theory and practical aspects of ML (e.g. how to debug, pipeline of a ML project)

Personally, I would have preferred if less starter code had been provided. I thought it would be useful to learn about how to create a machine learning application from scratch. It's also been many years since the course was first introduced. Would be great to do a renewal of it with more high-res graphics and state-of-the-art examples!

Thank you Prof Andrew Ng, I've learnt much from this course!

автор: Nikolay S

21 окт. 2016 г.

It would be easy to rate the course with anything below 5 stars for it being not enough formal and academic; for explicitly not requiring prior knowledge of basic calculus and algebra; for using MATLAB instead of actual industry standards.

But that would probably be unfair, because Andrew Ng's course is a brilliant introduction to Machine Learning. Soft, tolerant approach helps newbies to overcome initial feeling of being overwhelmed by ML algorithms and learn them while playing with lots of code provided with the course.

It is also definitely worth mentioning that prof. Ng not only explains problem settings and algorithms that are suited to solve them, but also shares many experience-driven hints for actually applying ML in practice.

All in all, this course will not make you a Data Scientist, but it definitely will help you grasp the basics and prepare you for more demanding education; or even for simpler actual practical tasks!

автор: AbdulSamad M Z

22 дек. 2018 г.

Machine Learning by Andrew Ng is one of the best courses I've ever taken - hands down.

The course is extremely well-organized and thoroughly covers the wide range of topics required to get a solid understanding of machine learning not only in theory, but also its applications and, more importantly, how to implement and debug it. The lessons cover a particular part of each chapter and are crystal clear. Although audio quality is not top-notch sometimes, the subtitles are there to help. The quizzes and assignments supplement your understanding by getting a hands-on experience on how to implement each concept and optimize your implementation. The assignment questions are crystal-clear in their requirements and contain additional code that visualize your work thus allowing you to focus on machine learning concepts and understand every teensy little concept related to them.

This course is worth every second and penny spent on it.

автор: Anton S

2 июня 2017 г.

This is a brilliant course. I hardly can express my experience in that short review. But I clearly feel that Andrew Ng is a very dedicated and talented teacher, as well as a great ML/CS professional. As a bit of PhD student myself I know also that mr. Ng is a real influencer in ML field, being the (co-)author of namely LDA, which is a fantastic idea. With mentioning this I just want to say that you hardly can find a better lecturer for ML. For the course, it is well-balanced, dense enough, with both in-depth and overview topics, acceptable complexity for programming assignments, and really holistic view on both research and applied aspects of the fascinating field of ML. I highly recommend everyone - whether you are a gonna-be ML engineer/researcher or just a curious CS/IT profi - take this course! It's high interdisciplinary and definitely worth learning, giving you a great insight into many issues of modern science.

автор: Stian R S

2 февр. 2017 г.

Excellent introduction to different methods in machine learning. I have some prior experience with machine learning, and although this is an introduction, it gave me a lot of good tips for implementation, debugging and workflow. It also gave me a deeper understanding of the different concepts in machine learning and when to apply the different methods. Questions during the lecture videos and quizzes afterwards keeps up your attention, and the programming assigments make you understand more deeply how to implement and apply the methods on real problems. The programming exercises are mostly pretty easy if you have some experience with programming and matrix/vector multiplication, and some of them are really funny to play with and apply on your own data or pictures afterwards. The lecturer, Andrew Ng, is also very good at explaining, and you never feel that he jumps over unclear details. I highly recommend this course!

автор: Paul P

21 сент. 2020 г.

Machine Learning was just the right amount of challenge for me, and really forced me to think both analytically in terms of what algorithms and programs are doing, and holistically/strategically, giving me some tools to anticipate common problems in machine learning system design. I do believe having a bit of background in mathematics and a beginner's grasp in programming helped me in the beginning, so if you don't have this knowledge, just be prepared to spend a bit more time on the material as well as brush up on some theoretical basics, such as linear algebra basics and sums/derivatives. The estimated time for completing each week of the course doesn't include the homework, which I found to be the most practical, useful, and challenging part. I would allocate 2x the estimated course time for each week if you include the homework assignments, unless you're already fluent in matrix multiplication and programming.