SEAnet: A Deep Learning Architecture for Data Series Similarity Search

TL;DR

SEAnet introduces a deep learning architecture that improves data series similarity search by learning effective summarizations, outperforming traditional SAX-based indexes especially on noisy or complex datasets.

Contribution

The paper presents SEAnet, a novel neural network architecture with new training strategies for data series summarization and similarity search, addressing limitations of existing methods.

Findings

SEAnet achieves higher accuracy in similarity search across diverse datasets.

DEA learned by SEAnet provides high-quality data series summaries.

SEAnet outperforms SAX-based indexes under challenging dataset conditions.

Abstract

A key operation for massive data series collection analysis is similarity search. According to recent studies, SAX-based indexes offer state-of-the-art performance for similarity search tasks. However, their performance lags under high-frequency, weakly correlated, excessively noisy, or other dataset-specific properties. In this work, we propose Deep Embedding Approximation (DEA), a novel family of data series summarization techniques based on deep neural networks. Moreover, we describe SEAnet, a novel architecture especially designed for learning DEA, that introduces the Sum of Squares preservation property into the deep network design. We further enhance SEAnet with SEAtrans encoder. Finally, we propose novel sampling strategies, SEAsam and SEAsamE, that allow SEAnet to effectively train on massive datasets. Comprehensive experiments on 7 diverse synthetic and real datasets verify the advantages of DEA learned using SEAnet in providing high-quality data series summarizations and similarity search results.