Not-so-simple linear regression with R

Рет қаралды 6,438

3 ай бұрын

Simple linear regression isn't so simple! We'll do a test/train split, build and diagnose a model, consider p-values, look for influential points, make predictions, compute root mean squared error, and more. This is a deep dive! If this vid helps you, please help me a tiny bit by mashing that 'like' button. For more #rstats joy, crush that 'subscribe' button!

Пікірлер: 15

@EquitableEquations 3 ай бұрын

You can find materials supporting this vid (and others) at github.com/equitable-equations/youtube.

@suprajm3117 8 күн бұрын

Very well explained, excellent teacher👌

@khalidmohamed6373 3 ай бұрын

Awesome ! clearly explained in a simplified way. Thanks, sir

@thiagobandini 3 ай бұрын

Initial split in content: linear regression building and evaluation AND training and validation sample sets 👏🏼👏🏼👏🏼

@erikvanhatalo7 3 ай бұрын

You are doing a fantastic job in this video! I refer my students to many of your videos Sir! Cheers from Sweden!

@EquitableEquations 3 ай бұрын

Cheers from Chicago!!

@munenedenis3377 3 ай бұрын

Please advise since am getting the following issue, I have already installed the tidymodels package "library(tidymodels) Error: package or namespace load failed for ‘tidymodels’ in loadNamespace(i, c(lib.loc, .libPaths()), versionCheck = vI[[i]]): namespace ‘rlang’ 1.1.1 is already loaded, but >= 1.1.3 is required In addition: Warning message: package ‘tidymodels’ was built under R version 4.3.3 "

@haraldurkarlsson1147 3 ай бұрын

It is hard to find a dataset of the appropriate values to showcase linear regression. I have looked through a bunch of the toy dataset supplied with R and some of the addons (Modeldata, CarData etc). You need data big enough to split. Any suggestions other than airquality?

@EquitableEquations 3 ай бұрын

Most packages come with good data sets. {ggplot2} and {modeldata} are good starting points. You can see all available data sets in loaded packages with `data()`. Good luck!

@haraldurkarlsson1147 25 күн бұрын

@@EquitableEquations I am aware of all the usual places. I was looking more for big datasets thar are off the beaten track. I have combed through Kaggle at no awail. I rarely find data in may own domain in the earth sciences (geochemical and/or isotopic data).

@user-dt9zw2wg9o 3 ай бұрын

Thank you i am vey aspired to be data scientist i took datacamp course what you recommend how much i should deep dive into statistic

@EquitableEquations 3 ай бұрын

Ideally the more the better, but time is finite and there are other important skills. It depends on the sort of data scientist you want to be. Good luck!