Predviđanje nuspojava lijekova pomoću odabranih metoda strojnog učenja temeljem javno dostupnih podataka o nuspojavama i neželjenim posljedicama primjene drugih lijekova

master's thesis

• Title (english): Prediction of drug side effects using selected machine learning methods based on publicly available data on side effects and adverse effects of drug use
• Author: Petra Žrvnar
• Mentor: Ljiljana Brkić (mentor)
• Committee member: Ljiljana Brkić (predsjednik povjerenstva)
• Committee member: Mirta Baranović (član povjerenstva)
• Committee member: Igor Mekterović (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2022-06-28, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Prije stavljanja lijeka na tržište, lijek ili cjepivo mora proći kroz sve faze kliničkog ispitivanja. Iako su klinička ispitivanja stroga i opsežna, postoji mogućnost da u ispitivanjima neće biti detektirane sve nuspojave lijeka. Nove nuspojave je moguće identificirati tek nakon puštanja lijeka u promet, prateći nuspojave u populaciji nakon korištenja lijeka. Ovakav pristup otkrivanju novih nuspojava je opasan, ali i skup, te je potrebno naći način detekcije novih nuspojava prije njihove pojave u populaciji. Kao rješenje ovog problema se nameće korištenje metoda strojnog učenja. U ovom radu je prikazan način korištenja odabranih metoda strojnog učenja za ranu detekciju nuspojava koje nisu otkrivene u fazi kliničkih ispitivanja.

Abstract (english)

Before a medication, or a certain vaccine is introduced, the medication in question undergoes a series of strict clinical trials. Even though clinical trials are strictly regulated, the possibility of an adverse reaction going undetected in the phase of clinical trials exists. Thus, these undetected reactions can only be identified after they appear in the general population. This way of detecting new, undiscovered adverse reactions is a very dangerous, but also a very costly method. The need to find other methods of detecting adverse reactions arises. A solution for this problem can be found in machine learning. This thesis describes selected methods of machine learning that can be used for early detection of undiscovered adverse reactions.

• Keywords: nuspojave; lijekovi; cjepiva; klinička ispitivanja; strojno učenje; stablo odlučivanja; L-BFGS algoritam; logistička regresija; binarna logistička regresija; regresija; binarna klasifikacija; ML.NET; farmakovigilancija; interakcija lijeka i cjepiva; utjecaj demografskih obilježja; VAERS; SIDER
• Keywords (english): adverse reaction; medication; drugs; vaccine; clinical trials; machine learning; decision tree; L-BFGS; algorithm; logistic regression; binary logistic regression; regression; binary classification; ML.NET; pharmacovigilance; drug and vaccine interaction; demographics; VAERS; SIDER
• Language: croatian
• URN:NBN: urn:nbn:hr:168:061265
• Study programme: Title: Computing; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-01-13 14:20:40