Logistic Regression Using Excel Video 2 - Predicting Heart Disease Probabilities

This is a follow up video to • Logistic Regression Us...
In this video I finish the model I started in the earlier video and create a confusion matrix then find the accuracy, prediction, and recall measures. I finish the presentation by creating a ROC (receiver operating characteristic) curve.
The link for the Excel file I used is provided below as a Google Sheets file. Do download it please open it, go to File- Download As - Excel
docs.google.co...

[@jupingert]

Thank you for sharing your knowledge. Both the videos put together are really really clear.

[@MelissaPompilius]

This video and its companion (Video 1) are very helpful and well-presented. Thank you for sharing!

[@tingey9291]

Thank you very much for sharing this in depth tutorial, it has proven to be invaluable to me. I wanted to ask, if possible, after determining the efficacy of the model, would you be willing to publish a tutorial detailing how this can be utilised in predictive practice? Your time and consideration would be appreciated.

[@tstan8121]

Extremely useful and enlightening. Thank you!

[@kaisong4529]

This video and previous one are great and helpful. They solved many questions in my mind. Thanks for sharing! Very nice!

[@Chroniom]

Thank you for the excellent explanations in the video! I have a small question: how could one find out the significance of the coefficients using excel?

[@mcanbolat]

Excellent question! It will be very different from the approach we take in regular regression. There is a resource here which may help:www.real-statistics.com/logistic-regression/significance-testing-logistic-regression-coefficients/ I may create a video on this myself.

[@humbertofrancamarcelo]

Thank you so much for your help with this video! Awesome work!

[@karthick2254]

Very good explanation, thank you for the great effort!!!

[@MOHITGUPTA-di4hc]

Great video. Thanks you How do I fit S curve on this? Any suggestions?

[@maheshsaivarma]

Thanks for such a great video, in my case the solver is not able to solve it, it is showing an error that one of the cells in the worksheet became an error value when solver tried certain values. Please help me what to do in this case

[@mcanbolat]

Is it a different dataset? Assuming no error in the model, sometimes the solver may stuck if a division by zero error occurs. You can start with arbitrary coefficents to possibly prevent that. You may have to try other software, this is just to explain how the model works. R or other software can easily give you the coefficients.

[@pranavsalvi8338]

Thank you Mustafa! Your video helped me.

[@ApoorwVikramDwivedi]

Using same data we can also find out R2 of the model but how do you find out p-value from this?

[@StellaJeruk-v4u]

thanks sir, but i have one question that how to test the model use the data test? cause in this video is only show us how to train a model.

[@mcanbolat]

You can use the logistic regression formula from the training data on a test data to see how it performs.

[@StellaJeruk-v4u]

@@mcanbolat Thank you very much for the reply & information, but i still don't know how, i'm newbie in machine learning, i'm sorry sir can you explain it to me, how it's work? Which one formula that i can use to test the data?

[@gembelningrat66]

is it the same for solving multinomial regression?

[@-MochMiftachurRifqiAlHusain]

Sorry sir, i have little question here about the evaluation model. This video explain us about regression and evalute with confusion matrix. How it works? Is it able to do regression evaluation using a confusion matrix? Why not use MSE, MAE, etc? Thank you sir

[@mcanbolat]

Hi, Logistic regression is a classification model and not a least squares regression model.The quality of the model is based on how good the classification is and it does this by maximizing the joint probability of success.

[@-MochMiftachurRifqiAlHusain]

@@mcanbolat ohh i see, thank you very much sir, i've been confused all day about that. I think the name is Logistic "Regression" that is regression model..

[@jeffreymun8221]

How do you decrease the sum of the log likelihoods more? Mine comes out to be -1800 to after the solver ran through it -1766..

[@mcanbolat]

Sometimes the solver does not do a good job if there are many variables. You can start with different initial values or try the evolutionary solver option in the menu. The best thing to do is to use a statistical software to get the coefficients. The objective of this video is to explain how the model works.

[@StevenDocherty.]

Would you be open to help with the development of a complex template?

[@tagalogmathclub6457]

Hi sir, to make sure, this is ordinal logistic regression right? This is not the multinomial logistic regression if I got it correctly.

[@mcanbolat]

Hi yes, multinomial means the dependent variable has more than two categories, not just two.

[@elenagoudarzvand6861]

Great job, well described , thanks

[@viveknandanwar8458]

thank you sir very helpful explanation

[@heriheryadi8817]

clear explanation

[@mateotorres8139]

so helpful!!

[@luyu4923]

Very helpful, thank you.

[@yalcnozturk4592]

Hocam çok teşekkürler, selamlar.

[@ElMakk_Med]

thank you