Voice Call Anomaly Detection (VCAD) is described herein to detect inconsistencies in patterns. VCAD is an anomaly detection system which is based on a long short-term memory (LSTM) algorithm and statistical methods. By detecting inconsistencies in patterns, the models described herein may detect and alert user of unusual voice service behavior that if not properly corrected can degrade, and possibly disrupt, the voice service. The statistical and machine learning methods used by VCAD are generic and may be used for solving other time-series problems when using other type of logs such as call logs, game logs, application usage logs, etc. The VCAD proactive, predictive capabilities allow customers to either eliminate the issue altogether, or turn costly, unplanned outages into controlled maintenance windows.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.