Hi Dr Rai, thank you for your teaching series. I have different threshold of correlation and VIF from yours, wondering which to follow? ---Mine--- Correlation > 0.8 (On Github I saw many people use it) VIF > 5 ( James et al. 2014) ---In this video--- Correlation > 0.5 (I feel that it is really low threshold) VIF > 10 (It is very high, I would rather prefer this threshold. Ha!) Thank you
,In one of the datasets for professor's salary , we have a factor variable of sex i.e Male and Female , with rule of n-1 factor variable as co-efficients , we observe Male coefficient is not significant . Does that mean Female co-efficient is also not significant?
![BLACK BAG - Official Trailer [HD] - Only in Theaters March 14](http://i.ytimg.com/vi/Du0Xp8WX_7I/mqdefault.jpg)
![Blox Fruits Dragon Rework Update [Full Stream]](http://i.ytimg.com/vi/EqDAp8udhm0/mqdefault.jpg)
Thank you so much, sir!
You are welcome!
Hi Dr Rai, thank you for your teaching series. I have different threshold of correlation and VIF from yours, wondering which to follow?
---Mine---
Correlation > 0.8 (On Github I saw many people use it)
VIF > 5 ( James et al. 2014)
---In this video---
Correlation > 0.5 (I feel that it is really low threshold)
VIF > 10 (It is very high, I would rather prefer this threshold. Ha!)
Thank you
People use either one as long as you can justify why.
hello professor, how do you address imbalance dataset when doing multiple regression, i.e a dataset that 40% of the data set contains NA,s
That's not imbalance, but it is missing data problem. Use this,
ruclips.net/video/An7nPLJ0fsg/видео.html
,In one of the datasets for professor's salary , we have a factor variable of sex i.e Male and Female , with rule of n-1 factor variable as co-efficients , we observe Male coefficient is not significant . Does that mean Female co-efficient is also not significant?
That's correct. It means sex as a variable is not significant.
@@bkrai Thanks for the reply ,looking forward to your live lectures
Next one will be tomorrow. See you soon.