thank you so much. your lecture helped me understand more on logistic regression. Quite easy to understand and make easier for me in my research analysis
Such a life-saver!!!! Thank you sir!!! Big help to me!! I was studying for my academic research paper / undergraduate thesis right now. The entry of pandemic in 2019 resulted to a halt in our semester and and so our lectures too. This was one of those we were not able to cover. But now, I have to self-study about it because I am going to need it for my research. Thank you thank you very much!! You’re a gift & a blessing!! You don’t know how much anxiety and worries were released from me. Praise God for your life and knowledge and heart to share knowledge to others
boy, this is humbling. I am sure it is me, but I have now looked at half a dozen places, and they all say the same thing regarding number 1 assumption for logistic regression. you said something about an example towards the end of the video. is that coming out or is it already out? I am not sure how to do this. meaning, in the linear case, you check the ACTUAL xs and ys. but correct me if I am wrong, I assume in the logistic case, you check the MODEL logs with the xs, correct ? just not sure how to "bring that home". thanks John
This is exemplary, well done. Quick question, for the linearity assumption, what about for x values that result in very high or very low probabilities? a logit function is relatively linear in the middle (say between probabilities of .2 to .8) but very steep at higher probabilities. In this case if you have x values that span the whole log odds function, wouldn't this naturally deviate from linearity? or am I missing something wrt what a glm logistic model is trying to accomplish? thanks!
I was wondering if I could get your advice on something related. If I were to do logistic regression on time series data and used time as a continuous predictor, and say I was measuring rates of disease that hovered around 50% but then dropped to
