Statistics 101: Nonlinear Regression, Introduction to Overfitting

In my latest Statistics 101 video we learn about the basics of overfitting, why complex models are not always the best, and about the balance between reducing error in our sample set while also being able to generalize to other data.
My playlist table of contents, Video Companion Guide PDF documents, and file downloads can be found on my website: www.bcfoltz.com
#statistics #regression #machinelearning

[@BrandonFoltz]

All video playlists available here: kzbin.infoplaylists Thank you and enjoy! 😄

[@jameslucas5590]

Enjoy your videos. A great refresher to what I've learned. Can you do one on how to read the statistic output? IE: The quadratic model shows that on average 63 cars sold for ~1.5 weeks on the job. With a selling range of 21-106 cars. I think I said that right.

[@SachinSingh-wr5yv]

Thank you sir...

[@davidboozer2410]

This video really highlights the importance of understanding the model you are trying to fit to your data. Polynomials are great because they're easy to use, analyze, integrate and differentiate, but they usually make horrible models for physical data... especially the higher order ones where ringing can be a problem. I really liked your perspective and how you explained the model is trying to "cater" to the individual data as opposed to be a good predictor of actual behavior. That's the first time I've heard someone describe polynomial fits like that and it's so true.

[@phowops]

What an amazing explanation! Love it!!

[@razhankers]

Thank you. Love your videos and examples with sample data

[@fredygonzalocarvajalserran1654]

As usual....Excellent...

[@azmaryaktermukthy7173]

very helpful.Thanks.

[@michalchik]

How are the P values for each order of term computed?

[@dineshv8761]

can you please do a series on Design of Experiments ?

[@waleros]

Thanks, Brandon!! very nice and well-explained video. HAve you thought about Generalized Non-Linear Models?

[@nazmusshakib5056]

Bradon Foltz thank you so much for your quality content. It's great. Please tell me is it good for machine learning and which playlist to follow first? I'm short in mentoring, anybody help?

[@jaygomez7037]

Hi Brandon! Thank you for this! Btw, what theme you are using to your graph in R?

[@wahidrahman3199]

Great job sir and i think ur presentation quit simple to understand :)

[@Shuffellove]

Hello, I have a long question! I am a master student doing Partial Least Squares as part of my project I dont quite understand how to test if my PLS models are significant and if they have enough power. Can you help me in that regard? Clara

[@MontahaAlriQa]

in multiple regression, do we fit one model for each DV? or 1 model in a high dimensional space? and what is the purpose of Multiple R measure?

[@glauberbarreto5135]

Top