Logistic Regression in R

[@wayarberry]

Excellent delivery. Just the right pace and the key points were clear. It sounds highly practiced! One of the things that seems to trip up a lot of you tube presentations is the presenter's fear of missing some obscure point and then later receiving criticism from a peer or "expert." They then launch into needless digressions which confuse the "zero day R person." This is an intro and you stuck to that mission. Great job.

[@BusinessScience]

Thanks a lot for the positive comments. Glad you enjoyed it.

[@rohitekka2674]

This is a brilliant example of end to end logistic regression in R, out there. Thanks!

[@BusinessScience]

Thanks so much. You are welcome!

[@AdilKhan-sh9fv]

While splitting the train vs test, is it possible for the train set to contain excessive rows for a particular year, thus skewing the model? Is there a function to balance the #of year types I'm both, train n test sets?

[@BusinessScience]

Yes, what we normally do is pick a stratification variable. In the resampling functions, there is a strata parameter. This keeps approximate proportions of that feature in both the train and test sets when resampling.

[@Sunny-China3]

Very informative video. Sir, kindly make video on poisson regression and and how to calculate euclidean distance of sites. Thanks

[@BusinessScience]

Thank you. I’ll look into this.

[@aaskyboi]

I'm getting the following error when I enter the code for prediction_class_test : Error in model.frame.default(Terms, newdata, na.action = na.action, xlev = object$xlevels) : factor fl has new levels c

[@RavinderRam]

awesome tutorial

[@BusinessScience]

Thank you!! =)

[@marcellberto2538]

Hey Matt, Came across this gem of a tutorial, thanks for sharing! A couple of questions: 1) When evaluating the AUC, you used results_tbl %>% roc_auc(year, .pred_1999). Is there a reason why you didn’t use .pred_2008 instead? 2) In the feature importance chart, what do the numbers along the x-axis represent? Your tutorials are great!

[@Rohan-Ahad]

This tutorial is great! I wonder if it is possible to run logistic regressions for multiple response variables against several explanatory variables in a single coding? I would like to run logistic regressions for multiple outcomes (4 dependent variables) and 4-5 explanatory variables from the same dataset. The conventional approach is time consuming that runs individual logistic regressions (for each response variable) and summarise the outputs in a table. Do you have any advise or tutorial please? Thank you.

[@BusinessScience]

Made a linear regression + map() tutorial that should help: kzbin.info/www/bejne/rZvCcnSpabt2bq8

[@Rohan-Ahad]

@@BusinessScience Thank you. I have run the glm model using the map function

[@saikoushik8770]

where can I get mpg data shown in tutorial?

[@BusinessScience]

If you sign up for our R-Tips, we provide all of the data and code files. The mpg data set actually comes from ggplot2 package. Here's a link to our R-Tips Newsletter: mailchi.mp/business-science/r-tips-newsletter

[@Talkinglife]

Nice

[@BusinessScience]

Thanks

[@slprojecthub]

Is business science and data science are same?

[@BusinessScience]

I am Business Science. Data Science is a field of analytical & statistical study. I teach people like you how to apply data science to business.