I have electricity price data in a matrix form of order 24*2192, where 24 are hours and 2192 are the number of days. In my data, there are outliers that greatly affect accurate forecasting, such that forecasting accuracy error is large. So, how to detect outliers and treat them in an effective way. One way to deal with such a problem study in literature is the moving window filter approach. Can someone guide how to apply this method to my data in R? please help that how I implement this method in R.
Regard
Ruey Tsay has done much in this area. See the attached Google search. Best wishes, David Booth BTW much of his work is in the z-library if you care to look there as well.
See these links:
https://www.researchgate.net/post/How_to_detect_and_treat_Outliers_in_Time_series_data
https://www.researchgate.net/post/How_do_I_remove_outliers_from_time_series_data
https://cxl.com/blog/outliers/
https://www.imsl.com/blog/pre-processing-data-time-series-analysis
https://stats.stackexchange.com/questions/69874/how-to-correct-outliers-once-detected-for-time-series-data-forecasting
The following links can help you:
https://www.r-bloggers.com/2016/12/outlier-detection-and-treatment-with-r/
https://stackoverflow.com/questions/743812/calculating-moving-average
https://www.storybench.org/how-to-calculate-a-rolling-average-in-r/#:~:text=To%20calculate%20a%20simple%20moving,from%20COVID%20in%20the%20US.
see the link
https://statsandr.com/blog/outliers-detection-in-r/
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis