Deep Learning on Attributed Sequences

TL;DR

This paper develops deep learning models tailored for attributed sequences, addressing the challenge of dependencies between sequence data and fixed attributes, and demonstrates significant performance improvements on real-world datasets.

Contribution

It introduces new deep learning solutions specifically designed for attributed sequences, a complex data type combining fixed attributes and variable-length sequences.

Findings

Models outperform state-of-the-art methods on real-world attributed sequence datasets.

Proposed solutions significantly improve task-specific performance.

Extensive experiments validate the effectiveness of the models.

Abstract

Recent research in feature learning has been extended to sequence data, where each instance consists of a sequence of heterogeneous items with a variable length. However, in many real-world applications, the data exists in the form of attributed sequences, which is composed of a set of fixed-size attributes and variable-length sequences with dependencies between them. In the attributed sequence context, feature learning remains challenging due to the dependencies between sequences and their associated attributes. In this dissertation, we focus on analyzing and building deep learning models for four new problems on attributed sequences. Our extensive experiments on real-world datasets demonstrate that the proposed solutions significantly improve the performance of each task over the state-of-the-art methods on attributed sequences.