Effect of Choosing Loss Function when Using T-batching for Representation Learning on Dynamic Networks

TL;DR

This paper investigates how different loss functions impact the training of dynamic network models using T-batching, proposing alternatives that improve performance on synthetic and real-world data.

Contribution

It introduces two novel loss functions tailored for T-batching in dynamic networks, addressing limitations of existing loss functions and enhancing model accuracy.

Findings

Proposed loss functions outperform original in synthetic networks

Achieved over 26.9% improvement in MRR on real-world data

Real-world tests show over 11.8% better Recall@10

Abstract

Representation learning methods have revolutionized machine learning on networks by converting discrete network structures into continuous domains. However, dynamic networks that evolve over time pose new challenges. To address this, dynamic representation learning methods have gained attention, offering benefits like reduced learning time and improved accuracy by utilizing temporal information. T-batching is a valuable technique for training dynamic network models that reduces training time while preserving vital conditions for accurate modeling. However, we have identified a limitation in the training loss function used with t-batching. Through mathematical analysis, we propose two alternative loss functions that overcome these issues, resulting in enhanced training performance. We extensively evaluate the proposed loss functions on synthetic and real-world dynamic networks. The results consistently demonstrate superior performance compared to the original loss function. Notably, in a real-world network characterized by diverse user interaction histories, the proposed loss functions achieved more than 26.9% enhancement in Mean Reciprocal Rank (MRR) and more than 11.8% improvement in Recall@10. These findings underscore the efficacy of the proposed loss functions in dynamic network modeling.