Machine Learning for Stress Detection from Electrodermal Activity: A Scoping Review

Abstract

Early detection of stress can prevent us from suffering from a long-term illness such as depression and anxiety. This article presents a scoping review of stress detection based on electrodermal activity (EDA) and machine learning (ML). From an initial set of 395 articles searched in six scientific databases, 58 were finally selected according to various criteria established. The scoping review has made it possible to analyse all the steps to which the EDA signals are subjected: acquisition, preprocessing, processing and feature extraction. Finally, all the ML techniques applied to the features of this signal have been studied for stress detection. It has been found that support vector machines and artificial neural networks stand out within the supervised learning methods given their high performance values. On the contrary, it has been evidenced that unsupervised learning is not very common in the detection of stress through EDA. This scoping review concludes that the use of EDA for the detection of arousal variation (and stress detection) is widely spread, with very good results in its prediction with the ML methods found during this review.

A Review on Machine and Deep Learning for Semiconductor Defect Classification in Scanning Electron Microscope Images

“…Therefore, the number of categories to classify the dataset is known. Once the system has learned to identify the different patterns, the classifier is able to assign each piece of data to its corresponding category [21]. There are several methods of supervised learning.…”

“…The goal is to obtain the highest separation between the data of the dataset. In order to achieve this goal, the initial dataset is split by means of binary divisions into branches along several iterations where the entropy is reduced [21,25]. The process ends up when the maximum tree depth is reached or a run-time cut-off is met [26].…”

“…The different samples of the cluster will be surrounding that point. The idea is to establish a number of clusters that cover the maximum number of samples while minimising the inertia (the distance between the cluster point and the different samples that belong to that cluster) [21]. During our search, one paper implementing the k-means algorithm to analyse SEM images was found [33].…”

“…A SOM or self-organising neural network is an unsupervised clustering method that allows to map the high-dimensional input data to a low-dimensional field (usually bi-dimensional) without losing the topological structure [35]. The maintenance of the topological structure is performed by applying a neighbourhood function [21]. Throughout our search, we also found one paper that implements an SOM method to perform the defect detection and classification task [36].…”

Analysing the Performance of Stress Detection Models on Consumer-Grade Wearable Devices

“…Many approaches to build an automatic emotion state/stress level discriminator using biometric data have been proposed, but most of them investigate the performance of the detection model using professional grade, highresolution devices in controlled laboratory settings [25,18,15]. Moreover, although the performance of stress detection models using Heart Rate (HR) and Heart Rate Variability (HRV) from wearable devices as well as data validation of HR signal and HRV are approved by many works [13,19,24,25,14,9], there is limited use of low-resolution (consumer-grade) EDA signals recorded from wearable devices and little is known on the resultant effect on the performance of stress detection [23,15,18,21,8,26]. Therefore, in this paper, we concentrate on studying and comparing different approaches of constructing stress detection models using low-resolution EDA signals.…”