I am trying to use anomaly detection to find the anomalies in my time series, and if I find it, I will replace it with my past values. I'm trying to do this because I want to create an upper and lower bound to replace those anomalies and by using the past values will help me to create this bound. Is there any guidance or example, where I can learn to do this? Thanks!
1
-
$\begingroup$ How does your data look like? (normal and anomalies) $\endgroup$– Jon NordbyDec 18, 2021 at 18:57
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
2
A rudimental mechanical approach would be to track moving average & its moving standard deviation for a predefined time window of your dataset. Any handful of points that temporarily breach over or under your 3 standard deviation value are outliers (assuming no process shift) & you could replace these by the moving average.
-
$\begingroup$ Could you give an example? Like a code example. I'm not sure how to write this in code. @eliangius $\endgroup$ Dec 12, 2021 at 23:47
-
1$\begingroup$ @user3002936 If you have the data in a panda dataframe, you can use
pandas.DataFrame.rollingto calculate rolling statistics and combine it withpandas.DataFrame.meanandpandas.DataFrame.stdto get the moving average and moving standard deviation. $\endgroup$ Dec 14, 2021 at 14:47