A few-shot learning approach with domain adaptation for personalized real-life stress detection in close relationships

TL;DR

This paper introduces a Siamese neural network-based metric learning approach for personalized stress detection in real-life settings, effectively addressing data scarcity and distribution mismatch issues in few-shot learning scenarios.

Contribution

It proposes a novel domain adaptation method using Wasserstein distance minimization within a Siamese network for personalized stress classification.

Findings

Improved stress classification accuracy in few-shot and one-shot settings.

Effective mitigation of distribution mismatch between target and non-target users.

Demonstrated applicability on real-life multimodal data from dating couples.

Abstract

We design a metric learning approach that aims to address computational challenges that yield from modeling human outcomes from ambulatory real-life data. The proposed metric learning is based on a Siamese neural network (SNN) that learns the relative difference between pairs of samples from a target user and non-target users, thus being able to address the scarcity of labelled data from the target. The SNN further minimizes the Wasserstein distance of the learned embeddings between target and non-target users, thus mitigating the distribution mismatch between the two. Finally, given the fact that the base rate of focal behaviors is different per user, the proposed method approximates the focal base rate based on labelled samples that lay closest to the target, based on which further minimizes the Wasserstein distance. Our method is exemplified for the purpose of hourly stress classification using real-life multimodal data from 72 dating couples. Results in few-shot and one-shot learning experiments indicate that proposed formulation benefits stress classification and can help mitigate the aforementioned challenges.