I was looking recently into weighted linear regression. Apparently, in the case that the residuals violate the homoscedasticity assumption, one thing you can do is to give weights to the values when calculating the least squares function. It seems to me that both the weighted linear regression and GLMs with the link function are trying to solve the same problem with different techniques. Is that so or I am missing something? Thanks!
Hi Dustin! Do you have a video where you explain Box cox transformation? would you recommend applying such transformation when the data is not normally distributed with multimodality?
Thanks! and I am confused :-) The function your call log function for me is a natural exponential function. Do you call it log because you need to use the log to solve it? or do I have the concepts confused?
Thank you to you both, very comprehensive and helpful.❤❤❤
I was looking recently into weighted linear regression. Apparently, in the case that the residuals violate the homoscedasticity assumption, one thing you can do is to give weights to the values when calculating the least squares function.
It seems to me that both the weighted linear regression and GLMs with the link function are trying to solve the same problem with different techniques. Is that so or I am missing something? Thanks!
really impressive, really
Hi Dustin! Do you have a video where you explain Box cox transformation? would you recommend applying such transformation when the data is not normally distributed with multimodality?
I found it! Here you apply box cox transformation using r:
th-cam.com/video/to3wgr6JuHE/w-d-xo.html
Thanks! and I am confused :-) The function your call log function for me is a natural exponential function. Do you call it log because you need to use the log to solve it? or do I have the concepts confused?
Probably. I'm not entirely consistent in my reference to logs/natural exponents.
Thank you!!
🤣🤣🤣 COOLEST STATS LESSONS. THANK YOU !!!