Gradual Fine-Tuning with Graph Routing for Multi-Source Unsupervised Domain Adaptation

TL;DR

This paper proposes a graph-based gradual fine-tuning framework for multi-source unsupervised domain adaptation, improving source domain selection efficiency and model accuracy on NLP tasks.

Contribution

It introduces a novel graph routing approach with error bounds for optimal training order in multi-source domain adaptation.

Findings

Achieved 2.3% accuracy improvement on NLI task.

Attained 3.9% accuracy boost on diverse Sentiment Analysis subset.

Developed lightweight strategies for source domain selection.

Abstract

Multi-source unsupervised domain adaptation aims to leverage labeled data from multiple source domains for training a machine learning model to generalize well on a target domain without labels. Source domain selection plays a crucial role in determining the model's performance. It relies on the similarities amongst source and target domains. Nonetheless, existing work for source domain selection often involves heavyweight computational procedures, especially when dealing with numerous source domains and the need to identify the best ones from them. In this paper, we introduce a framework for gradual fine tuning (GFT) of machine learning models on multiple source domains. We represent multiple source domains as an undirected weighted graph. We then give a new generalization error bound for GFT along any path within the graph, which is used to determine the optimal path corresponding to the optimal training order. With this formulation, we introduce three lightweight graph-routing strategies which tend to minimize the error bound. Our best strategy improves $2.3\%$ of accuracy over the state-of-the-art on Natural Language Inference (NLI) task and achieves competitive performance on Sentiment Analysis (SA) task, especially a $3.9\%$ improvement on a more diverse subset of data we use for SA.