Conformal Kernel Expected Similarity for Anomaly Detection in Time-Series data

The problem of anomaly detection arises in many practical applications. Currently itis highly important to be able to detect outliers in data streams, as recent years have seen a rapidgrowth in the amount of such data. Only a few techniques are applicable to real-time data and evenfewer could provide an interpretable anomaly score. Probabilistic interpretation of the anomalyscore could allow an analyst to choose the anomaly threshold based on the desired false alarm rate,which is highly important in a number of real-life applications. We propose a modification of theEXPoSE algorithm for anomaly detection in time series data, which produces a probabilistic scoreof abnormality. The proposed algorithm is developed within the framework of conformal anomalydetection and utilizes the expected similarity as a measure of non-conformity.