Primjena strojnog učenja u predviđanju cijena avionskih letova

master's thesis

• Title (english): Application of Machine Learning in Prediction of Airfare Prices
• Author: Matej Lopotar
• Mentor: Tomislav Burić (mentor)
• Committee member: Tomislav Burić (predsjednik povjerenstva)
• Committee member: Josipa Pina Milišić (član povjerenstva)
• Committee member: Lenka Mihoković (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2024-07-09, Croatia

Abstract

Cijene avionskih karata variraju prema mnogim faktorima kao što su dan leta, gradovi polaska i dolaska, trajanje leta, broj dana do leta, broj presjedanja i aviokompanija kojom se leti. Zrakoplovne kompanije određuju cijene na temelju algoritama koje su razvile. Testiranje i analiza provedeni su na skupu podataka koji opisuje njemačke letove.
Cilj istraživanja je na temelju dostupnih podataka o letovima predvidjeti cijenu avionskih karata koristeći različite regresijske algoritme koji se temelje na stablima odlučivanja.
Prije samog predviđanja podaci su detaljno analizirani i stršeće vrijednosti su uklonjene. Implementirana su tri regresijska algoritma: RandomForestRegressor, ExtraTreesRegressor i DecisionTreeRegressor. Rezultati su opisani pomoću sljedećih metrika: Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Squared Error (RMSE), R² koeficijent determinacije i Mean Absolute Percentage Error (MAPE).

Abstract (english)

Airline ticket prices vary according to many factors such as flight day, departure and arrival cities, flight duration, number of days until the flight, number of stopovers, and the airline company. Airline companies determine prices based on algorithms they have developed. Testing and analysis were conducted on a dataset describing German flights.
The aim of the research is to predict airline ticket prices based on available flight data using various regression algorithms based on decision trees. Before the prediction itself, the data was analyzed and outliers were removed. Three regression algorithms were implemented: RandomForestRegressor, ExtraTreesRegressor, and DecisionTreeRegressor. The results are described using the following metrics: Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Squared Error (RMSE), R² coefficient of determination, and Mean Absolute Percentage Error (MAPE).

• Keywords: Stabla odlučivanja; Strojno učenje; Analiza podataka; Predviđanje cijena
• Keywords (english): Decision Trees; Machine Learning; Data Analysis; Price prediction
• Language: croatian
• URN:NBN: urn:nbn:hr:168:777466
• 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: 2024-10-12 21:21:37