Anomaly-aware summary statistic from data batches

TL;DR

This paper introduces a parallelized approach to the New Physics Learning Machine (NPLM) for large-scale data analysis, enhancing computational efficiency and sensitivity in detecting subtle deviations in collider data, with applications in offline and streaming scenarios.

Contribution

It proposes a batch-wise parallel NPLM method that improves resource efficiency and sensitivity, enabling anomaly detection in large or streaming datasets.

Findings

Outperforms simple sum of batch tests in sensitivity.

Can match or surpass full data test performance.

Enables anomaly-aware summary statistics in streaming data.

Abstract

Signal-agnostic data exploration based on machine learning could unveil very subtle statistical deviations of collider data from the expected Standard Model of particle physics. The beneficial impact of a large training sample on machine learning solutions motivates the exploration of increasingly large and inclusive samples of acquired data with resource efficient computational methods. In this work we consider the New Physics Learning Machine (NPLM), a multivariate goodness-of-fit test built on the Neyman-Pearson maximum-likelihood-ratio construction, and we address the problem of testing large size samples under computational and storage resource constraints. We propose to perform parallel NPLM routines over batches of the data, and to combine them by locally aggregating over the data-to-reference density ratios learnt by each batch. The resulting data hypothesis defining the likelihood-ratio test is thus shared over the batches, and complies with the assumption that the expected rate of new physical processes is time invariant. We show that this method outperforms the simple sum of the independent tests run over the batches, and can recover, or even surpass, the sensitivity of the single test run over the full data. Beside the significant advantage for the offline application of NPLM to large size samples, the proposed approach offers new prospects toward the use of NPLM to construct anomaly-aware summary statistics in quasi-online data streaming scenarios.