Anomaly detection methods have been developed for streamed data; these methods can be applied to the telecommunications data. Anomalies within telecommunications data are sometimes consequences of critical incidents; therefore, fast optimal decision-making after anomalies have been detected within BT is important to ensure critical national infrastructure is maintained.

A decision-making algorithm utilising a pre-trained optimisation is not ideal because methods applying non-adaptive decision-making policies have been found to be unable to learn how to make optimal decisions in the streaming data setting where reward distributions may be non-static; therefore, a bandit approach would be more suitable for this problem. The decision-making algorithms we develop will relax unsuitable assumptions commonly made in multi-armed bandit policies to synthesise multi-armed bandit techniques with well-established anomaly detection methods. The novel decision-making algorithms we will develop will be self-optimising and adaptive. Furthermore, the algorithm will give feedback to the anomaly detection method to improve the accuracy and delay of detection.