A high jump from the previous courses. It was not well organized. Besides, it is hard to find peers to review the final project, hence to get the certificate.

This course was the most challenging one among all courses in specialization. I wish there were more explonation of how we get smth from smth and not like " and it appears to be equal .." and so on.

I also had to watch Ben Lambert's bayesian course on YouTube to understand the material of the second week,because Prof. David Banks was not good enough in explanation.

Assistant Prof. Mine Çetinkaya-Rundel has an amazing teaching skills.

Prof. Merise Clyde is good in explonations and I understand that she tried to present a very complicated material in a simple way, but as I have already mentioned above, I wish there were more explonations of casuality of the formulas with examples and Intuition that stands behind these formulas like in Ben Lambert's videos.

Many thanks for such an amazing experience.

In my opinion this is the most difficult course in this specialization. There is a lot of new concepts to master and, although Bayesian reasoning is more natural, most of us has been conditioned to think as a frequentist, not as a Bayesian. However, the course offered a glimpse on how Bayesian approach can deal certain issues where frequentist approaches fail and that is the most important lesson one can take home from this course.

It would be very helpful if the teachers provide us an indication of a good book on Bayesian Statistic that is friendly to people that are not mathematically oriented (like myself).

It was nice learning all the distribution functions and Bayesian statistics. However, I have one suggestion: When going through equations, it's better to dive a little deeper into them, or at least go through a few steps of derivation, rather than just show them on the screen. For example, in 'Bayesian Regression' when introducing 'conjugate bivariant normal-gamma distribution, it was directly given three correlations on the screen: (1) alpha | sigma^2 ~ N(a0, sigma^2 S_alpha, (2) beta | sigma^2 ~ N(b0, sigma^2 S_beta), (3) 1/(sigma^2) ~ G(mu_0/2, mu_0 sigma^2/2. There are many terms in the equation. It would be more learner friendly if one can at least go through what term corresponding to what. Or if time is a constraint one can at least show some reasonable reference, so that learners can search for papers. I had to do quite an amount of googling to get through these things.

Too many formulas... More examples would be nice.

Honestly, I find it very hard to recommend this course to anyone. First of all, let's note that the course covers quite more advanced topics than the previous 3 courses in the specialization, so some extra difficulty is to be expected. However, the main problem I encountered was with the video lectures themselves, particularly during the 3rd and 4th week of the course. The instructor does a very rushed job at explaining everything, constantly giving us tons of information and jargon that is not previously mentioned, and even the examples fail to give us insight at what we need to do and why.

Other than the lectures, no external material is given to help us decipher what the professors are saying, other than a few Wikipedia links. I'm not saying that each course should be accompanied by an e-book, but honestly, if I wanted to learn about Bayesian Statistics from Wikipedia I could have well skipped this class.

The main reason I'm giving 2 stars to the course instead of 1 is the Labs and the Quizzes. Even though they could use some polishing too, especially the final Lab, they are indeed very helpful and do a much better job at clarifying the concepts presented.

All in all, I feel that if you want to learn about Bayesian Statistics you should look for another course, and/or save your money and get yourselves a good textbook.

This course could have been so much more, but sadly, it wasn't. The first 3 courses in this series were absolutely brilliant. However, this course falls way short on expectations. In comparison to the very thoughtful explanation and pace of the first 3 courses, this course introduced a giant leap in self learning requirements and overall challenges. It seems to be very out of touch with what it touts to offer - an introductory course in Bayesian Statistics to learners. It questioned us on topics which hadn't been introduced yet. This is the only course in the series where I didn't learn any statistics, and just tried to out-game the quizzes and assignments.

This course is way below the quality of the three previous ones. I is really a shame. I suggest it to be totally redesigned. The text book is not very well written. Besides grammatical errors it also has inconsistencies with what is taught in the lectures. As for the lectures they seem to have been done in a hurry, and many important aspects are not clearly explained. Coursera and Duke should think of splitting this course in two and leave the rush and lack of detail, that can be seen in this one, behind. I really believe this course is below Duke's standards and is a black spot in the Specialization.

Bayesian statistics is hard, I get it. This is another reason not to throw a huge amount of concepts on students, with no explanations, nor any sense. I had to study Bayesian Statistics by myself, and out of this course. Please correct this issue.

I was following the courses very well until I got to this. I had trouble understanding almost all parts of this one:/

The material was rushed and did not provide enough examples for a beginner to understand.

This course is absolute garbage. Halfway through week 3 nothing makes sense anymore. They switch up the teachers on you and start throwing massive formulas in your face with huge chunks of R code that only use functions made specifically for this course that act like black boxes where I have no idea what they're doing. The instructional content is all super old and doesn't even make sense. Their R package doesn't even work anymore, they have all kinds of grammatical errors and typos in the videos and written materials, and none of the lessons feel like they build on information from previous lessons.

I'm really interested in Bayesian statistics and have a strong stats background and this course has only made me more confused than before I started. I've never been so frustrated with a course (online or in person) in my entire life. Don't waste your time or money.

Unlike the 3 prior courses in this specialization, Bayesian Statistics is not taught well at all. I had the following issues with the course: (1) The course book is full of typos making it frustrating to read (2) You are expected to understand lots of statistical jargon that is not introduced in prior courses or even explained well in the course that it is being used in (3) The lessons are not contextualized into how these techniques are applied to the real world.

It's disappointing to go through 3 courses that were taught extremely well by Professor Mine Çetinkaya-Rundel only to end it on a course like this which seemed to be half-baked and speedily covered. There may be other courses on Bayesian Statistics on Coursera that teach this increasingly important field better.

Taking this course was honestly a really upsetting experience.

Starting in week 3 the instructor you are used to gets swapped out for someone who simply cannot communicate concepts or ideas. Your only options will be to either start gaming the quizzes or find external resources to figure out what is going on. Even in the latter case, the course expects you to use their own custom functions in R for the assignments. They no longer maintain the course or their R library and some of the function were incomplete to begin with. Anyway, even when you fight your way through it and then wait days for the peer reviews to come in, you will try to finally move on to the capstone project just to find it locked until some arbitrary start date set one to two months in the future.

Absolutely useless and not at "beginner" level as compared to the rest of specialization. This one part requires reading almost as much as all other parts in specialization combined. Videos=textbook read in front of camera. There was not made clear, why Bayesian approach is more useful than frequentialist approach in real world statistical analysis.

This course requires immediate review. It is incompatible with the others of this specialization. It is not intuitive, it relies heavily on dense mathematical formulas with no time for practice or memorization. The material presents errors and one of the R studios had bugs. It should be an specialization on its own.

I want to give 0 ratings. The worst course I have seen so far in Coursera. Horrible planning, horrible execution and makes no sense. Totally disappointed by the style of course design and delivery

The professors tried to put too much material into very short videos.

The result is that most of the material are explained unclearly.

they didn't tell that this course didn't have a homework, reading, or practice problems to do. Ended up s

Very interesting and formative. It starts from the basics (Bayes' theorem) but then it goes beyond the usual conjugate models such as Beta-binomial and Gamma-Poisson. Bayesian Linear Regression, Bayes Factors, Bayesian Model Averaging and a brief introduction to MCMC are provided. This really put me in the position of applying Bayesian Statistics to some real world application: the final test case is a good illustration. The only minus is that the part on Bayesian Hypothesis Testing (in particular Bartlett's and Lindley's paradoxes) is a bit rushed up, and not as clear as the rest of the course. All in all, a really good course, I'm glad I followed it.

I beta-tested this course and it was an amazing experience. The instructors are super engaging and upbeat, despite having to delivery a very complex subject like Bayesian statistics. The quizzes are very rich, with a nice balance of practice quizzes and graded quizzes. The practice quizzes offer very helpful explanations that can help to reinforce understanding before doing the graded quiz. Last but not least, the final project description is super exciting and thorough. I highly look forward to completing this course and get a deep introduction to Bayesian statistics.

The instructors teach good. It is good to do the lab exercise after watching the video to apply it. The course cover wide range of topics. The treatment of priors is serious. It also good to know about the simulation. Final project will have opportunity for some critical thinking about some statistical problems. In review peer projects, I found out what others do better than me, and that different learners might think and treat problems differently.

I appreciate that the course materials contained a lot of practises and readings. Clearly the course organisers did not want to compromise the volume and difficulty of materials for popularity. The materials were concise and clear. A tangible sign of progress is that I can now understand some recent studies using Bayesian methods. It has given me a good start of learning more advanced Bayesian techniques.

this course is one of the best course in Coursera.

I learned a lot about Bayesain statistics, although the content is difficult, but it is presented in the simplest way possible, so in order to complete this course, you must adhere to what the supervisors say about this course.

Outstanding material. It may be the hardest level compared to the rest specialization course, since Bayes indeed have high technical level detail. But it was worth it. Great course and detailed from the instructors.