You should probably do a seminar to our professors on how to teach this. I'm a graduate student in Data Science and it took me 2 semesters to pass the bayesian statistics class. Stopped going to class because I couldn't understand anything, watched the series from this video, went to exams, and passed it with an A-. So thank you. This is the simplest way to understand Bayesian Statistics and very helpful even at the graduate level.
As far as I understand, the likelihood distribution is just the distribution of probability densities of x=2 over all possible parameters, which is p for binomial distribution. It is maxed at 0.2, since MLE of the binomial distribution is simply p. You probably manage to understood long ago, but some might find my explanation useful.
You should probably do a seminar to our professors on how to teach this. I'm a graduate student in Data Science and it took me 2 semesters to pass the bayesian statistics class. Stopped going to class because I couldn't understand anything, watched the series from this video, went to exams, and passed it with an A-. So thank you. This is the simplest way to understand Bayesian Statistics and very helpful even at the graduate level.
I fully agree. My teacher does this bottom up approach where he dives deep into the mathematics without any intuition/examples/concepts or whatever.
This.
This is an absolutely fantastic demonstration. Thanks for this! You can't beat a great visualization in making stats things like this crystal clear!
It is so amusing to watch this video, everything seem clear and intuitive. You are a true hero
What a marvellous presentation! Thank you Ben.
You videos are so so informative, I wish I could subscribe twice!
Great explanation, but should the X axis for likelihood be from 0 to 10, instead of 0 to 100, same as for prior and posterior?
As far as I understand, the likelihood distribution is just the distribution of probability densities of x=2 over all possible parameters, which is p for binomial distribution. It is maxed at 0.2, since MLE of the binomial distribution is simply p. You probably manage to understood long ago, but some might find my explanation useful.
May I ask how do you make the plot? Matlab or else?
well it says 'wolfram mathematica' student edition on top...
You could use langage R (free with the ide Rstudio, more appropriate for statistiques and plus), Python etc...
so useful
brilliant