Відео

finally someone explains it in a way that makes sense! thanks so much!

Very very helpful

thank you for your insightful videos

How to hit 5000.69 subscribers

This is cool man, I am doing genomics research currently and I have to implement multivariate generalized Bayesian linear regression to find how genetic factors simultaneously influence traits. Your research, though in a different field, is very helpful as a reference and interesting too!!

I'm totally failing this ;-;

why were we sampling from the prior distribution for posterior prediction? Seems like an obvious error.

Ma'am, can I get the lecture note pdf?

Great explanation. Has anyone ever thought of using these ideas for a language model? I could have continuous learning built in, due to the Bayesian Approach?

What book do you use?

Please which textbook are you referring to in the video?

what are the prerequisites for this course

Thank you for sharing those videos!

Thank you for your video! Your explaination is very clear and I've learned much from it. Please make more videos.

Hi Prof, i am not sure why " has the probability changed between 1992 and 1993" got interpreted mathematically into Pr(P1 < P2). It could also have been interpreted as Pr(P2 < P1) which would have a different result. The data samples are independent from each other (ie. data from 1992 is independent from 1993 and vice versa) hence, in my mind, the ordering should not matter.

Thannk you, excellent video!

but what the case for the difference between two dependent uniform random variables ? would you kindly make a case vedio for that

ok

excelente ejemplo/ejercicio! muchas gracias

谢谢老师 确实帮助比较大

Great video!

Okay.... Okay.... Okay

Thanks dia for reaching out to our understanding. Bit is there anyway I can access you slides in pdf format?

How do I do it for three iid uniform random variables?

which slides are you using? I can not find it in your github repo, the one with the restaurant one

Will ruin folks

ok? ok? ok? ok? ok?

When cdf is nondecreasing function is it right continuous??

Yes, Chinese explaining math. We can be 100% sure that it is correct.

A random variable itself is a set of values. These values can be interpreted as data points. E[X] is the weighted average (= the outcome that we most likely expect). Variance Var[X] is the degree of spread in the set of data points around E[X]. It shows the amount of variance among the data points. The larger the variance, the „fatter“ the distribution (= the graph is spread further from the middle point E[X]). Covariance Cov(X,Y) inspects how the values of two random variables X,Y correspond to each other. If Cov(X,Y) is large, then X and Y have correlating high values (= the values are high at the same points). For example: Studying more correlates to higher grades. But if Cov(X,Y) is negative, it shows that the high values of X correspond with the low values Y. For example: If it rains a lot, less people go outside.

There is a simple derivation for the expectation of X, s.t E[X] = n*p A binomial distribution Bin(n,p) is just the sum of n INDEPENDENT Bernoulli Distributions Y with probability p. As we know, E[Y] = p. Thus E[X] = n * E[Y] = n*p

Thank you so much you just made this so much easier ❤

The indicator function Z for an event A has ONLY two values. If x € A => Z(x) = 1 If x !€ A => Z(x) = 0 Since the sample space is just a partition of A and its complement A^c, we have: sample space = A u A^c Thus, E[Z] = 1*P[A] + 0*P[A^c] = 1*P[A] + 0*(1-P[A]) = P[A]

P[X = „Zero ex-girlfriends for CS majors“] = 1.0

{X <= b} = {X <= a} u {a < X <= b} (Union of disjunctive sets, use sigma-additivity of P): P[X <= b] = P[X <= a] + P[a < X <= b] ==> P[a < X <= b] = P[X <= b] - P[ X <= a] = F(b) - F(a)

Great tutorial. Please post more videos about DP

This course has solved so many questions I have for probability and I am feeling so much more confident about the coming up final! Thank you so much for your excellent explanation and for sharing those videos!

Thank you so much it helps a lot!

is there a text book this class is based upon ?

Thank u for explaining in detail!

Thanks

谢谢胡老师开放这门课程，对我帮助很大

谢谢胡老师开源这门课程，对研究生转到统计的我帮助很大

there is a little error in the writing of 5 choose 2 (you actually wrote what would be 5 choose 3 )

can you solve last question

Prof Hu. Very nicely explained. I have been following your videos and could make some decent progress. I have a question though; will the constants always get cancelled (immaterial of the choice of the prior)? If they don't, then looks like I will need to calculate the constant and will not have the luxury to drop them. Please advise.

Hey, thank you so much for uploading all these videos as I find those are very informative and helpful for clarifying my confusion. I was wondering how you approached the question that one of your students asked in the video (i.e., someone believes that every day out of the seven days are equally likely, how can we give a prior reflecting that belief for p itself). Specifically, what are possible values that p could take in that case? Many thanks ahead of time!

I want to say there is no comment but the content you've posted is high-value and much appreciated.

probability of saying 'okay' in a sentence i.e. P('okay') > 0.7 🤣🤣

really love your voice