What is it about?
This paper focuses on the binary classification of the emotion of fear, based on the physiological data and subjective responses stored in the DEAP dataset. We performed a mapping between the discrete and dimensional emotional information considering the participants’ ratings and extracted a substantial set of 40 types of features from the physiological data, which represented the input to various machine learning algorithms—Decision Trees, k-Nearest Neighbors, Support Vector Machine and artificial networks—accompanied by dimensionality reduction, feature selection and the tuning of the most relevant hyperparameters, boosting classification accuracy.
Featured Image
Why is it important?
The methodology we approached included tackling different situations, such as resolving the problem of having an imbalanced dataset through data augmentation, reducing overfitting, computing various metrics in order to obtain the most reliable classification scores and applying the Local Interpretable Model-Agnostic Explanations method for interpretation and for explaining predictions in a human-understandable manner. The results show that fear can be predicted very well (accuracies ranging from 91.7% using Gradient Boosting Trees to 93.5% using dimensionality reduction and Support Vector Machine) by extracting the most relevant features from the physiological data and by searching for the best parameters which maximize the machine learning algorithms’ classification scores.
Perspectives
The technology application scenario involves a future development of a wearable device that uses these algorithms for clinical applicability in the treatment of anxiety by VR immersion therapy. Because the use of VR equipment does not allow an easy collection of EEG data or interpretation of facial expressions, we chose EDA and HR as the only viable options in these conditions. As future directions of research, we also plan to apply the same feature extraction, classification and feature selection methods on data originating from other databases (such as MAHNOB [33]) and on data recorded from subjects who will play a virtual reality game dedicated to phobia therapy. We have developed and are currently developing various games for acrophobia [55], ophiophobia [56], claustrophobia [57], pyrophobia [58] and fear of public speaking therapy [59]. The best classification algorithms will be used to automatically estimate the in-game intensity of fear and to adjust the level of exposure to various threatening stimuli
Livia Petrescu
University of Bucharest
Read the Original
This page is a summary of: Machine Learning Methods for Fear Classification Based on Physiological Features, Sensors, July 2021, MDPI AG,
DOI: 10.3390/s21134519.
You can read the full text:
Resources
Contributors
The following have contributed to this page
