Zurück zu Bayessche Statistik

3.9

592 Bewertungen

•

184 Bewertungen

This course describes Bayesian statistics, in which one's inferences about parameters or hypotheses are updated as evidence accumulates. You will learn to use Bayes’ rule to transform prior probabilities into posterior probabilities, and be introduced to the underlying theory and perspective of the Bayesian paradigm. The course will apply Bayesian methods to several practical problems, to show end-to-end Bayesian analyses that move from framing the question to building models to eliciting prior probabilities to implementing in R (free statistical software) the final posterior distribution. Additionally, the course will introduce credible regions, Bayesian comparisons of means and proportions, Bayesian regression and inference using multiple models, and discussion of Bayesian prediction.
We assume learners in this course have background knowledge equivalent to what is covered in the earlier three courses in this specialization: "Introduction to Probability and Data," "Inferential Statistics," and "Linear Regression and Modeling."...

Sep 21, 2017

Great course. Difficult to apprehend sometimes as the Frequentist paradigm is learned first but once you get it, it is really amazing to see the believe update in action with data.

Apr 10, 2018

I like this course a lot. Explanations are clear and much of the (unnecessarily heavyweight) maths is glossed over. I particularly liked the sections on Bayesian model selection.

Filtern nach:

von Jaime R

•Nov 08, 2018

Theorethical backdrop is a bit excessive on an R focused course

von Maurits v d M

•Aug 22, 2016

I had a lot of fun during this course, but I think it is simply too short to present all the topics in sufficient detail. Furthermore, I took this course without doing the prior courses in the specialization, and there were a couple of moments when I really thought previous knowledge from a different course was required.

I think for the most part the lecturers did a great job in explaining the materials in the course. The lectures themselves were also well structured, and the topics followed each other in a logical order. I would have loved to spend more time on modeling techniques and Markov Chain Monte Carlo.

von 陈昊

•Nov 15, 2017

Harder than former courses but great!

von José L E N

•Jan 03, 2017

Theis course is substantially more difficult than the three first ones, and the material is scarce. However, I must admit that this is one of the courses I have ever learnt the most

von KALYESUBULA M

•Jun 03, 2017

Learnt a lot. Though the subject material was hard to grasp first hand, it is good that instructor was readily available to help us through.

von zhaokai

•Aug 16, 2016

I hope we can have access to the slides, and this can save attendees a lot of time, because I think after we finished watching the video we can skip relevant slides when we come across problems in doing exercise.

von Ángela D C

•Jun 12, 2018

Week 3 was too much information too soon, but week 4 was great again like the other courses in this specialisation. Learned so much, thanks!

von Emmanouil K

•Aug 16, 2017

This is a very interesting topic. Lectures in weeks 3 and 4 could use some work.

von Chen

•Jul 07, 2017

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.

von Kian B

•Jul 29, 2016

The section about Beta-Binomial Conjugate is taught very fast and unless the student is quite familiar with Beta and Gamma distributions, it makes it very difficult to follow the course.

von Vicken A

•Dec 29, 2016

Bayesian stats is a broad topic. Learners would benefit from more material.

von Hanyu Z

•Dec 08, 2016

The material is good. However, there is no support from the instructors to answer our questions in the discussion forum.

von George G R

•May 06, 2017

The classes are good.

von sohini m

•Oct 27, 2017

It was nice

von Stanley R C

•Jan 30, 2018

The instructors have great expertise, but this course is pretty difficult for a Bayesian newbie. Additional study guides would be helpful (especially week 4).

von José M C

•Mar 22, 2017

Good content but sometimes it gets confusing.

von Adam A

•Aug 25, 2017

An interesting and challenging course, would be better with more real examples and explanation as some of the material felt rushed

von Jae S P

•Jul 18, 2017

This is one of many good courses that one can get a glimpse of Bayesian statistics though it lacks of thorough explanation of mathematical background and reading materials of any kind.

von Mark F C

•Jun 21, 2018

It was a good course, though I would include more coursework and exercises in R to assist with comprehending a difficult subject. Overall, good course for something that's difficult to teach.

von Ravichandran V

•Aug 06, 2019

Its really hard for me to follow this specific course, its as if I am reading a summary of a novel rather than a novel, ideally this course should be broken into two courses and made into two five week courses. I may need to take additional courses or read some books to get a clear understanding.

In the previous three courses the open stats book helped a lot, however, the online content for this course is difficult to follow as well.

von Lalu P L

•Jun 02, 2019

The course could have been more comprehensive and less verbose. It had so much content in a tiny course. Content should be less and more comprehensive.

von Uira d M

•Jul 03, 2019

The course is well structured but the span of topics is large and the complexity great. Maybe an extended version with more explanations and demonstrations of the equations would be better for understanding the whole concept of bayesian statistics, specially inference.

von Niels R

•Jul 06, 2019

This course through the material too fast. The content should have been spread out over two courses in my opinion.

von Malolan S

•Sep 10, 2019

A bit more depth in explaining conjugacy in priors and posteriors will be very helpful. A possible way would be to have more example illustrations.

von Pouya Z

•Sep 26, 2019

The course was great and really informative. Particularly, it was interesting to get to work with BAS and statr packages that were developed, essentially, by the instructors. I, however, think that from decision loss functions onward, the course suddenly became way more complex. The normal conjugate families were not discussed on the previous lab, and I believe deserve to be emphasized with an example before heading to regression and reference priors. However, the notes were quite helpful. All and all, it was a great course.

Coursera arbeitet mit erstklassigen Universitäten und Organisationen zusammen, um Online-Kurse anzubieten und dadurch universellen Zugriff auf die weltweit beste Ausbildung zu ermöglichen.