Explaining Probability Distributions

A guide to how to understand probability distributions in the context of statistics

Stay updated with the channel and some stuff I make!

[rishipatel]

The place where you link Probability with Statistics is really good..
It's really helpful to make a picture about the things you deal with in your daily work.

[Unaimend]

I love your way of introducing concepts. I am a Comp Sci. student currently getting into medicine and those videos are insanely helpful. Looking forward to your more advanced videos

[pipertripp]

This was great! Having done some self-study on statistics in the past year, these videos are great for re-stimulating all the connections and also seeing some familiar concepts in a new light.

[andrews]

I just took a stats course in my Data Science Masters, and this video was a perfect summation of it. We also used R to visualize the distributions, albeit we didn't use ggplot (which looks nicer and is better IMO). Great job!

[itwasthesame]

I am still training to prepare to apply in Very Normal Company, this video really helps

[anthonymorford]

Good thing it's Sunday. Professional relaxers only. We can never be fired.

[subramanyanvishwanath]

Grateful to you ;as a grad student very helpful for me as a refresher video to go through

[ag]

Great explanation! This really helped me put all these concepts together.

[FTFP]

I think “T” is used because it stands for “Test Statistic”. That is, the reason we’d be interested in a statistic is because of its distribution which would be related to figuring out how interesting the value of it is under our distributional assumptions.

That, or it stands for the “t” in statistics since “s” is already used for standard deviation and sometimes sum.

[thunderbrain_]

Great video, thanks for your work! really clear and polish explanations

[rafaelcalderon]

so good thank u for inspiring students to learn this

[jankpeters]

Nice video and example. On the r example part I have one question though. Why did you use "runif(2, min=1, max=6) %>% ceiling", it will never give 2 or 3 as a result because it rounds up. Wouldn't "runif(2, min=0, max=6) %>% ceiling" be better for the double dice roll.

[bcs]

To complement your discussion on probability vs statistics, one can think of Probability as knowing the data generating process and studying the properties of the data generated by it, and of Statistics as knowing the data and studying the properties of the data generating process that could have generated it

[backtoGodhead]

It was helpful, thanks for creating it.

[walterreuther]

In 5:21 I believe there is an error.
If you plug in half-open intervals (-∞, x] to a probability distribution, you get the cdf.

This distinction is actually important because there are discrete RVs, which have a probability mass function (pmf), and continuous RVs, which have a probability density function (pdf), but - and this is where it is relevant - there are also mixed RVs, that are described by neither.
This is a pain to deal with if you want to write a general form for anything - for instance an expected value. Still, with a little measure theory (and I mean really little, I hardly know any of it myself...) it is possible to generalise...

[alexrosas]

Just a quick note, on 4:27 you said that all random variables have a pdf, but that is not always true. The easiest example is a cdf which is not differentiable, then the pdf can not exist. Much more complicated examples can be constructed, but the previous is a quick one.

[calebmerritt]

I’m not an expert in R, but wouldn’t you need a uniform distribution from 0-6 if you’re rounding up?

[santiagodm]

I found your channel and I feel like I find a gold mine. Hey, could recommend some books or resources for a beginner in stats .

[elijahsagaran]

Isn't a statistic denoted by "T" because it's also referred to as Test-statistic?

[CharlesLampman]

Robot brain happy. Data input saved to hard drive.