MACHINE LEARNING AND EMOTION RECOGNITION - VALIDATION OF AN ALGORITHM - AND PERSPECTIVES OF USE

Aug 5, 2026

The study examines how emotional regulation shapes optimal educational trajectories by validating an algorithm that identifies a subjective emotional state from facial-expression recognition. The proposed Artificial Intelligence model relies on the two-dimensional Arousal–Valence framework, producing an emotional output from observed cues. Validation is pursued through a careful correlation analysis between the algorithm’s results and complementary neurophysiological monitoring collected via biofeedback, linking affective inference with physiological evidence.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages14

Reading Time35 min

AccessFree to download

Introduction

Emotions in learning contexts

Affective computing and emotion-based adaptive learning systems

MACHINE LEARNING AND EMOTION RECOGNITION: VALIDATION OF AN ALGORITHM AND PERSPECTIVES OF USE MACHINE LEARNING E RICONOSCIMENTO DELLE EMOZIONI: VALIDAZIONE DI UN ALGORITMO E PROSPETTIVE D'USO Emanuele Marsico Pegaso Telematic University (UNIPEGASO) [emanuele.marisco@unipegaso.it](emanuele.marisco@unipegaso.it) Double Blind Peer Review Citazione Marsico E., Barbieri U., Piceci L (2023), Machine learning and emotion recognition: validation of an algorithm and perspectives of use, Giornale Italiano di Educazione alla Salute, Sport e Didattica Inclusiva - Italian Journal of Health Education, Sports and Inclusive Didactics. Anno 7, V 2. Supplemento Edizioni Universitarie Romane Doi: [https://doi.org/10.32043/gsd.v7i2.967](https://doi.org/10.32043/gsd.v7i2.967) [Copyright notice:](Copyright notice:) © 2023 this is an open access, peer-reviewed article published by Open Journal System and distributed under the terms of the Creative Commons Attribution 4.0 International, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. [gsdjournal.it](gsdjournal.it) ISSN: 2532-3296 ISBN: 978-88-6022-479-8 ABSTRACT The literature emphasises the role of emotional regulation in determining optimal educational trajectories. Therefore, we propose to validate a useful algorithm for the identification of subjective emotional state through the recognition of facial expressions. Such an Artificial Intelligence (AI) algorithm is based on the twodimensional Arousal-Valence model. The aim will be achieved through a careful analysis of the correlation between the output of the algorithm and the complementary neurophysiological monitoring performed by means of biofeedback. La letteratura sottolinea il ruolo della regolazione emotiva nel determinare traiettorie educative ottimali. Pertanto, si propone di validare un algoritmo utile per l'identificazione dello stato emotivosoggettivo attraverso il riconoscimento delle espressioni facciali. Tale algoritmo di Intelligenza Artificiale (AI) si basa sul modello bidimensionale di Arousal-Valence. Lo scopo sarà raggiunto attraverso un'attenta analisi della correlazione tra l'output dell'algoritmo e il monitoraggio neurofisiologico complementareeffettuato tramite biofeedback. KEYWORDS Emotion Recogniton, Machine Learning, Biofeedback, Artificial Intelligence Riconoscimento delle emozioni, Apprendimento automatico, Biofeedback, Intelligenza Artificiale Received 11/09/2023 Accepted 26/09/2023 Published 26/09/2023 Umberto Barbieri Niccolò Cusano University [umberto.barbieri03@gmail.com](umberto.barbieri03@gmail.com) Luigi Piceci Niccolò Cusano University [luigi.piceci@unicusano.it](luigi.piceci@unicusano.it) Introduction1 Emotions are psychophysiological and relational processes that influence individuals' cognitive and behavioral functions. In academic settings, emotions are involved in modulating attention, motivation, memory, and the transfer of acquired knowledge (Mayer, 2019). Furthermore, they can either facilitate or hinder learning depending on how they are recognized, managed, and integrated with thinking processes (Jerath R, Beveridge C. Respiratory Rhythm, 2020; Usán Supervía, P., & Quílez Robres, A., 2021) . For example, Pekrun (2021) explored emotions related to reading and learning from texts, which are primary sources of information and knowledge. The author emphasized the importance of emotions in text comprehension, memory formation, as well as in generating new ideas and problem-solving. Particularly, by activating various brain areas involved in cognitive processes such as the amygdala, hippocampus, prefrontal cortex, and dopaminergic system, emotions modulate the encoding, consolidation, and retrieval of information in long-term memory (Tyng, C. M., Amin, H. U., Saad, M. N.,& Malik, A.S., 2017) . Moreover, the literature highlights a positive correlation between experiencing adapt

FAQs

What problem does the document address?

It addresses the need to validate an algorithm capable of identifying subjective emotional states from facial expression recognition for educational purposes.

What model does the AI algorithm use for emotion identification?

It is based on the two-dimensional Arousal–Valence model.

How is the algorithm validated in the study?

By analyzing the correlation between the algorithm’s output and complementary neurophysiological monitoring performed through biofeedback.

Aug 5, 2026

Upload by Olivia Brown