Predviđanje ishoda Formule 1 pomoću strojnog učenja

master's thesis

• Title (english): Predicting Formula 1 Outcomes with Machine Learning
• Author: Josipa Kaselj
• Mentor: Marina Bagić Babac (mentor)
• Committee member: Marina Bagić Babac (predsjednik povjerenstva)
• Committee member: Mirjana Domazet-Lošo (član povjerenstva)
• Committee member: Marko Horvat (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2023-07-10, Croatia

Abstract

Fokus ovog diplomskog rada bio je razviti učinkovito i djelotvorno algoritamsko rješenje koje koristi tehnike strojnog učenja za točno predviđanje budućih ishoda utrke Formule 1. Analizom povijesnih podataka koji su uključivali statistiku performansi vozača i tima, uz informacije o vremenu, rezultati kvalifikacija i rezultati utrka korišteni su za treniranje modela. Prethodno obrađivanje podataka uključivalo je uklanjanje nedosljednosti u skupovima podataka i prilagodbu samog formata podataka. Implementacijom različitih metoda kao što su linearna regresija, regresor gradijentnog pojačanja, regresor slučajne šume i regresor višeslojnog perceptrona, fokus je bio na predviđanju završnih pozicija svakog od vozača. Drugi cilj bio je predvidjeti samo pobjednika utrke korištenjem modela poput logističke regresije, klasifikatora slučajne šume, klasifikatora k susjeda i klasifikatora višeslojnog perceptrona. Nakon učenja svakog od ovih modela, za njegovu svrhu, testirana je točnost i ukupna izvedba kako bi se provjerila sposobnost modela da generalizira nove podatke.

Abstract (english)

In this Master’s Thesis project main focus was to develop an efficient and effective
algorithmic solution utilizing Machine Learning techniques to make accurate predictions about Formula 1 racing future outcomes. By analyzing historical data that included driver and team performance statistics, alongside weather information, qualifying results, and races results were used to train the model. Data pre-processing took
many forms including cleaning up inconsistencies in datasets and adapting data formats. Implementing various methods such as Linear Regression, Gradient Boosting
Regressor, Random Forest Regressor, and Multi-Layer Perceptron Regressor models,
the focus was on forecasting all driver’s finishing positions. The second goal was to
predict only the race winner by leveraging Logistic Regression, Random forest Classifier, k−Neighbors Classifier, and Multi-Layer Perceptron Classifier models. After
training each of this model, for its purpose, its accuracy and overall performance was
tested in order to check model’s ability to generalize to new data.

• Keywords: Python; umjetna inteligencija; strojno učenje; Formula 1; regresija; klasifikacija; Gradient Boosting Regressor; Multi-layer Perceptron Classifier
• Keywords (english): Python; Artificial Intelligence; Machine Learning; Formula 1; regression; classification; Gradient Boosting Regressor; Multi-layer Perceptron Classifier
• Language: english
• URN:NBN: urn:nbn:hr:168:525617
• 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: 2023-10-11 09:16:07