Napovedovanje športnih rezultatov v nogometu s pomočjo strojnega učenja

diplomsko delo

Jakob Simičak (Author), Iztok Fister (Mentor), Grega Vrbančič (Co-mentor)

Abstract

V zaključnem delu smo se osredotočili na statistiko pričakovanih zadetkov v nogometu. S programom Figma smo ustvarili izgled programa, ki smo ga poimenovali Footstat. Program Footstat uporablja podatke podjetja Statsbomb, ki je med vodilnimi podjetji v zbiranju in obdelovanju nogometnih podatkov. Z uporabo njihovega API-ja smo lahko dostopali do podatkov preko Python knjižnice. Omejili smo se na brezplačne podatke, ki jih je podjetje namenilo za raziskovalne in študijske namene. Nato smo ustvarili metriko s pomočjo logistične regresije, ki smo jo implementirali s pomočjo knjižnice za obdelavo podatkov v Pythonu Scikit-learn. Končni rezultat je postal program Footstat, ki je s pomočjo metrike izračunal pričakovane zadetke za izbrane tekme glede na omejene podatke podjetja Statsbomb . Izračunane pričakovane zadetke smo na koncu primerjali z dejanskimi zadetki na tekmah in analizirali morebitna odstopanja.

Keywords

pričakovani zadetki;nogomet;logistična regresija;Statsbomb;Footstat;diplomske naloge;

Data

Language:	Slovenian
Year of publishing:	2024
Typology:	2.11 - Undergraduate Thesis
Organization:	UM FERI - Faculty of Electrical Engineering and Computer Science
Publisher:	[J. Simičak]
UDC:	004.85:796.332.093(043.2)
COBISS:	221065475
Views:	0
Downloads:	27
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	English
Secondary title:	Expected results in football with the help of machine learning
Secondary abstract:	In my final project we focused on calculating Expected goals statistics in Football. With Figma, a software for designing prototypes, we created how we wanted our software, which we named Footstat, to look. Footstat uses data from a company named Statsbomb, one of the leading companies in analysing football data. With the use of their API, we could access the data through their Python library. We limited ourselves to using their free to use data, which were “gifted” for use in academic and research purposes. Then we made our metric with the help of logistic regression, which we implemented with the help of the Python library scikit-learn. This all resulted in the final software Footstat, which, with the help of our metric, calculated the expected goals for the chosen matches in our limited data from the Statsbomb company. At the end, the calculated expected goals were compared to the actual goals scored in matches.
Secondary keywords:	expected goals;football;logistic regression;Footstat;Statsbomb;
Type (COBISS):	Bachelor thesis/paper
Thesis comment:	Univ. v Mariboru, Fak. za elektrotehniko, računalništvo in informatiko, Informatika in tehnologije komuniciranja
Pages:	1 spletni vir (1 datoteka PDF (X, [47] f.))
ID:	24794565