Klasifikacija afektivnih poremećaja metodama strojnog učenja

master's thesis

• Title (english): Classification of Affective Disorders Using Machine Learning Methods
• Author: Eda Jovičić
• Mentor: Mario Cifrek (mentor)
• Committee member: Mario Cifrek (predsjednik povjerenstva)
• Committee member: Igor Krois (član povjerenstva)
• Committee member: Željka Lučev Vasić (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2021-07-08, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Electrical Engineering

Abstract

Afektivni poremećaji su psihički poremećaji raspoloženja u koje spada povratni depresivni poremećaj. Elektroencefalografija je postupak snimanja električne aktivnosti mozga postavljanjem elektroda na površinu glave. Strojno učenje se bavi razvojem metoda za što točnije i učinkovitije učenje i generalizaciju podataka. U ovom radu, korišten je skup podataka u kojem se nalaze snimke EEG-a zdravih ispitanika i onih s dijagnozom povratnog depresivnog poremećaja. Na skupu signala EEG-a izvršena je... More predobradba i obradba s ciljem uklanjanja šuma i izdvajanja značajki. Izdvojena su dva seta podataka: set izdvojenih značajki i set značajki bipolarnog signala. Korištenjem metoda strojnog učenja, setovi značajki klasificirani su u dvije klase: „zdravi“ i „depresivni“. Za klasifikaciju korišteni su plitki modeli i konvolucijska neuronska mreža, te su rezultati klasifikacije prikazani u radu. Less

Abstract (english)

Affective disorders are psychiatric mood disorders that include recurrent depressive disorder. Electroencephalography is the process of recording the electric brain activity by placing electrodes on the surface of the head. Machine learning develops methods for more accurate and efficient learning and data generalization. In this paper, a data set containing EEG recordings of healthy subjects and those diagnosed with recurrent depressive disorder was used. The EEG signal was preprocessed to... More remove noise and processed to extract features. Two sets of data were created: a set of extracted features and a set of bipolar signal features. Using machine learning methods, feature sets were classified into two classes: “healthy” and “depressive”. Different types of shallow learning methods and a convolutional neural network were used for classification, and the results are presented in this paper. Less

• Keywords: afektivni poremećaji; elektroencefalografija (EEG); strojno učenje; konvolucijske neuronske mreže (CNN)
• Keywords (english): affective disorders; electroencephalography (EEG); machine learning; convolutional neural networks (CNN)
• Language: croatian
• URN:NBN: urn:nbn:hr:168:108069
• Study programme: Title: Electrical Engineering and Information Technology; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka elektrotehnike i informacijske tehnologije (magistar/magistra inženjer/inženjerka elektrotehnike i informacijske tehnologije)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-01-13 10:07:27