Statistics 101: Linear Regression, Confidence Bands - Part 1 of 2

[@Tech2188]

longer is better. this is perfect for something im actually working on at work right now(fun with manufacturing and statistics). I'm really looking forward to part 2! Love your videos! keep up the good work! oh and first i guess.

[@KirillBezzubkine]

God bless you. I m saying that for a 100th time. U even arranged all vids into playlists.

[@jayjayf9699]

I’m watching these videos for my actuary exams, and I have to learn linear regression in my notes it spits out a formulae for mean and predicted response and I have no intuition for what the understanding is between them as it’s so ambiguous, hopefully once I’m done watching your videos I can understand the difference better

[@tomashorych394]

Thanks! nice and clear explanation. I used the ci option randomly in my thesis and right now, evening before the defense I realized I can't interpret it :D so thank you!

[@ZwickySarkozy]

At 14.13 it looks like the video was interrupted. Like there were a few slides still to come but then it cuts to the 'Great Courses' promo instead and then the video abruptly ends? There's a fraction of a second where the 'Final Points' slide appears but it quickly disappears and isn't discussed...

[@meokha]

Thank you for the great work and helping tons of people understand stats!

[@cococnk388]

Thanks Sir....so great a teacher

[@pragun1993]

I understand that since there can be many samples, our expected tip value for a given bill value is not just one number you get by putting values in the equation of y = b0+b1x. It will be an estimate for all the different values you get for different samples. But you can have only one mean for those different expected values of different samples. How can you have a number of different means for one input value of bill amount. You can have a distribution for different expected values of y for a given x(bill amount) because of different samples, but how can you have different mean values and their distribution? it will be just one mean expected value for all those different expected values for different samples, right?

[@xuchuan6401]

For a given set of samples, you get a y^ for a specific x, this y^ is an estimator for E(Y) at this point x, so the y^ is inherently a specific realization of the distribution of E(Y). By having another set of samples, you get a different y^. In other words, these y^ are all means themselves, which together constitute E(Y).

[@بناءالمجتمع-خ5ع]

This is G-R-E-A-T! If I could, I would have tipped you on paypal. Keep the good work

[@BrandonFoltz]

Thank you! I do have a GoFundMe ❤️

[@Muben4889]

Wonderfull man, your work is benefigial

[@jamesmangini4064]

Great videos on this playlist. Thank you. Do you have any videos how to use regression for segmentation?

[@jianli1692]

Thanks. What if there is only one data point with $64, how to get the distribution?

[@rossgai8630]

Thanks from Ross and Cassie

[@ycbarton]

thank you very much!

[@jamesmangini4064]

Do you have any videos how to use regression for market segmentation?

[@astaconteazamaiputin]

Did you check out the whole channel ? Did you also search for it on his website ?

[@MC-rs6jp]

Hi Brandon, thank you so much for these videos... so clear and helpful! @13:57 you discuss the tip amount and the CI lower and upper. When the bill is $34, the CI includes zero, does this mean that we must accept the null hypothesis at this bill amount?

[@sethhhh]

regresi tdk bersifat deterministik, tip sebesar $8,54 hanyalah salah satu dari banyak tip yg mungkin, so berapa tip rata rata untuk makanan sebesar $64?

[@gracege3836]

what distribution are you assuming?

[@BrandonFoltz]

Hi Grace! Assumption is normality, but it's pretty robust with respect to violations. Transformations are also OK and common if normality is violated in an extreme degree.

[@tatianaletourneau602]

thank u sir !

[@quochungto8329]

thank you so much

[@maxmusti8101]

Thank you so much :-) !

[@crystalbelot5885]

I miss the positive reinforcement at the start. It was nice to hear 😅

[@crystalbelot5885]

Thank you for breaking everything down visually 🙏

[@plekkchand]

"thinking in your head"?

[@panagiotisgoulas8539]

Brandon the notation is actually weird. I mean look at the slide on 09:35. Phrased like this is like saying constructing a ci for the estimate, although this is not what you are trying to say as explained earlier. The ^ symbol as far as I have seen is used for estimates, I prefer the notation of μ^Υ |X* for the estimate and μΥ |X* the parameter that you actually make the ci for.

[@BrandonFoltz]

Hi! As I say in the video, different authors, practitioners,instructors, etc. use different notation. I use the one I am familiar with. Sorry for the confusion. Hopefully someday everyone will settle on one way but I'm not holding my breath :)