5. Logistic Regression - Generating Logistic estimates using Excel and Log Loss

In this video we cover the intuition behind log loss and demonstrate the formula in an excel spreadsheet using a sample of variables on the Charles Book Club case study.
We then further demonstrate how to generate a simple Logistic Model using Excel Solver which minimizes the Log Loss over the dataset. The regression equation so generated is contrasted with the regression estimates using R - demonstrating that they are the same. The objective is to explain log loss in a practical example to the viewer.

[@suriebc6779]

It's made my day..I have struggled a lot to find working exaple like this.. Thank you so much..

[@UltraPokemonwalkthru]

Fantastic video explaining Logistic Regression and log loss. I have never seen anyone who is so fast with both Excel and R ! Kudos!

[@kadhirn4792]

Long live Man. Explained wonderfully. Never thought of doing it in excel. Saved my day. Kudos!

[@mohammedtaher3717]

Very well explained Thank You!

[@ravindarmadishetty736]

Excellent video Karan.....first time heard about logloss in my experience. Thank you. Very useful video

[@somdubey5436]

really really helpful and beautifully explained. Loved the way it was explained through both R and Excel.

[@balajinatarajan1051]

Excellent video to understand each step

[@pascaltorvic6246]

Great Video

[@sreevidyasankarasubramania733]

Thank you. Should I adopt the same method for 5-point Likert scale too. please Let me know.

[@SESHUNITR]

Good one

[@jayantapanigrahi4920]

You make it so simple. It’s just what Iookjng for. Also what is this solver?

[@LearnanalyticsIn]

Its an excel add in for optimization problems - free to use

[@NewYearResolution2024]

When i use the LogLoss formula in my excel file, it shows #NUM! . What is this error? Kindly advice

[@LearnanalyticsIn]

Formula is not written correctly

[@NewYearResolution2024]

Cell A - "Years with Company" Cell B - "Recency" Cell C - "Frequency" Cell D - "Monetory" Cell E - "Response" (Dependent Variable) " Values are 1 or 0" Cell F - Z Cell G - Probability Cell H - LogLoss LogLoss Formula = E4*LN(G4)+(1-E4)*LN(1-G4) Probability Formula = EXP(F4)/(1+EXP(F4)) Kindly advice whether the formula is right.

[@NewYearResolution2024]

Kindly allow me to send you the file. I am preparing for interviews. If I figure out how to do this, I can talk about it in my interviews. Kindly provide your email ID or please send a test mail to [e-mail].

[@karansarao2584]

I would suggest moving into R or SAS - rather than replicating in Excel. As long as you understand the concept - using actual procedures on a true statistical computing platform would be more useful for your preparation

[@bhalokhabaar7242]

try explaining in a way that the novice category of the viewers can also understand.

[@amitshrivastava4727]

Showing prob > 1