20.500.12556/DKUM-98614 Uporaba podatkovnega rudarjenja v košarki The use of data mining in basketball Namen diplomskega dela je bil raziskati uporabo podatkovnega rudarjenja in naprednih analitičnih metod v košarki s poudarkom na vrednotenju individualne uspešnosti igralcev. Raziskava se osredotoča na analizo naprednih statističnih kazalnikov in njihovo primerjavo z osnovnimi statističnimi podatki. V okviru diplomskega dela so bili analizirani podatki zadnjih treh sezon lige NBA in identificirani ključni statistični kazalniki, ki vplivajo na uspešnost igralcev. Igralci so bili razvrščeni v različne skupine glede na njihovo vlogo v ekipi. Analizirane so bile razlike med uspešnimi in manj uspešnimi igralci. Pri tem so bile uporabljene metode strojnega učenja, kot so metoda k-najbližjih sosedov, logistična regresija in naključni gozd. Rezultati raziskave kažejo, da napredni statistični kazalniki omogočajo bistveno natančnejšo analizo in napovedovanje uspešnosti igralcev v primerjavi z osnovnimi statističnimi podatki. Modeli, zgrajeni na podlagi naprednih statistik, dosegajo višjo stopnjo natančnosti, kar potrjuje njihovo uporabnost v sodobni športni analitiki. Raziskava potrjuje, da podatkovno podprt pristop predstavlja pomembno orodje za razumevanje košarke, ki lahko bistveno prispeva k boljšemu odločanju v športnem okolju. This thesis aimed to explore the application of data mining and advanced analytical methods in basketball, with a focus on evaluating individual player performance. The research emphasizes the use of advanced statistical metrics and their comparison with traditional statistics. The analysis is based on data from the last three NBA seasons, where key statistical indicators influencing player performance were identified. Players were classified into different roles within a team, and differences between high-performing and low-performing players were examined. Additionally, machine learning methods, including k-Nearest Neighbors (kNN), Logistic Regression, and Random Forest, were applied to evaluate their predictive capabilities. The results show that advanced statistical metrics provide significantly more accurate insights into player performance compared to traditional statistics. Models built on advanced data achieved higher prediction accuracy, confirming their effectiveness in modern sports analytics. The research demonstrates that a data-driven approach represents a valuable tool for understanding basketball performance and can significantly support decision-making processes within teams and sports organizations. podatkovno rudarjenje košarka napredni statistični kazalniki strojno učenje analiza podatkov uspešnost igralcev data mining basketball advanced statistics machine learning data analysis players' performance features true true false Slovenski jezik Angleški jezik Diplomsko delo/naloga 2026-06-24 09:57:53 2026-08-04 21:44:59 2026-08-05 03:06:15 0000-00-00 00:00:00 2026 0 Maribor 0 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 286921475 VS_Markovic_Filip_2026.pdf VS_Markovic_Filip_2026.pdf 1 1FEE77567949CBF2A5662D112C622ED8 56c264ef42b58e04f2e88a2702b42453c30e70922c18d1bff20efa7dc667d38b e7bf1f82-6fa2-11f1-868d-00155d580e34 https://dk.um.si/Dokument.php?lang=slv&id=207532 Fakulteta za organizacijske vede 0 0 0