- Using Prophet for timeseries anomalies
- Uses XGBoost to perform regression on top-10 features for predicting DL Throughput speeds
- Performs Daily Degradation of DL_THROUGHPUT
- Performs Weekly Degradation of DL_THROUGHPUT
- Trains Prophet model weekly
rajeevratan84 / lte-kpi-anomaly-detection Goto Github PK
View Code? Open in Web Editor NEWLTE KPI Anomaly detection using time series forecasting, regression modeling, and correlation differences in KPI. Also, daily and weekly KPI degradation reports generated.
License: MIT License