I have a heavy tailed data which is heteroscedastic in nature. I need to apply some ML algorithms on this data that are good for normal distribution. So I want to transform the data. I tried Box cox transformation, square root and log transformation but it does not help much and the data still retains its heavy tailed and heteroscedastic nature. What are some of the other variance stabilizing that can be of help here.

  • $\begingroup$ good question. Can you tell more regarding your method and data? $\endgroup$ – Peter Jul 31 '19 at 21:10
  • $\begingroup$ @Peter I am analyzing time series data. I analyzed heteroscedasticity using bptest method in R. But before and after transformation, the p values remains less than 0.05. QQ plot of the data also remains almost intact. $\endgroup$ – shaifali Gupta Jul 31 '19 at 21:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.