Di Bernardo, Alessandro (2024) Exploring the potential of quantum machine learning to improve analysis and classification of EEG signals: a comparative study with machine learning. [Tesi di dottorato]
![[thumbnail of DiBernardo_Alessandro_37.pdf]](https://www.fedoa.unina.it/style/images/fileicons/other.png "DiBernardo_Alessandro_37.pdf") |
Documento PDF
DiBernardo_Alessandro_37.pdf Visibile a [TBR] Amministratori dell'archivio Download (7MB) | Richiedi una copia |
| Tipologia del documento: | Tesi di dottorato |
|---|---|
| Lingua: | English |
| Titolo: | Exploring the potential of quantum machine learning to improve analysis and classification of EEG signals: a comparative study with machine learning |
| Autori: | Autore Email Di Bernardo, Alessandro alessandro.dibernardo@unina.it |
| Data: | 11 Dicembre 2024 |
| Numero di pagine: | 121 |
| Istituzione: | Università degli Studi di Napoli Federico II |
| Dipartimento: | Ingegneria Elettrica e delle Tecnologie dell'Informazione |
| Dottorato: | Information technology and electrical engineering |
| Ciclo di dottorato: | 37 |
| Coordinatore del Corso di dottorato: | nome email Russo, Stefano stefano.russo@unina.it |
| Tutor: | nome email Angrisani, Leopoldo [non definito] |
| Data: | 11 Dicembre 2024 |
| Numero di pagine: | 121 |
| Parole chiave: | quantum technology; metrology field; quantum machine learning; quantum computing; EEG dataset |
| Settori scientifico-disciplinari del MIUR: | Area 09 - Ingegneria industriale e dell'informazione > ING-INF/05 - Sistemi di elaborazione delle informazioni Area 09 - Ingegneria industriale e dell'informazione > ING-INF/06 - Bioingegneria elettronica e informatica Area 09 - Ingegneria industriale e dell'informazione > ING-INF/07 - Misure elettriche e elettroniche |
| Informazioni aggiuntive: | Candidato appartenente al ciclo 37. |
| Depositato il: | 28 Dic 2024 10:42 |
| Ultima modifica: | 12 Ago 2026 05:37 |
| URI: | https://www.fedoa.unina.it/id/eprint/16432 |
Abstract
Quantum technologies have attracted some attention from the scientific community in recent years, as promising in various application contexts such as artificial intelligence. From the perspective of quantum computation, there is a clear distinction between the bit, which can assume the values 0 or 1, and the qubit, which has the capacity to assume both values simultaneously. This distinction has significant implications for computational power. The objective of this study is to estimate the current potential of quantum computing, with a particular focus on quantum machine learning in terms of measurement. To this end, a decision-making prototype was devised using a dataset of EEG signals provided by the research team and contrasted with a classical machine learning approach, with an evaluation of its generalisation and discrimination capabilities. The implementation phase required a special effort, as there were a number of aspects to be chosen such as the development environment, framework, libraries and variables to be able to develop the classification algorithm. During the development process, there were some intermediate stages which showed not only the possibility of comparing a classical and a quantum approach, but also a hybrid approach distinguishing hybrid quantum machine learning and full quantum machine learning. The latter showed, from the results collected, that there are problems with generalisation and thus highlighting overfitting phenomenon. This preliminary experimental investigation showed that, although in some computational aspects QML seems very interesting and promising, there are a number of aspects such as the difficulty in generalising that highlight a premature state compared to classical machine learning.
Downloads
Downloads per month over past year
Actions (login required)
 |
Modifica documento |
