I want to schedule data learning for a source and alert me more accurately when the data gets closer to zero and this behavior is not normal. I am currently using Forecast time series with a learning time of 150 days backwards but it generates false alerts, any suggestions to adapt my model?
