Details

Napovedovanje šahovske uspešnosti z uporabo fizioloških parametrov
ID Drab, Žiga (Author), ID Bratko, Ivan (Mentor) More about this mentor... This link opens in a new window, ID Luštrek, Mitja (Comentor)

.pdfPDF - Presentation file, Download (2,71 MB)
MD5: 18E77F54DC8ABB8637FDF31AAF7B4528

Abstract
Stres se odraža v merljivih fizioloških spremembah in lahko vpliva na odločanje pri kognitivno zahtevnih nalogah. Cilj raziskave je bil ugotoviti, ali fiziološki parametri omogočajo napovedovanje šahovske uspešnosti in šahovskih dogodkov ter izboljšajo napovedi na podlagi šahovskih značilk. Podatke, zajete med uradnim šahovskim turnirjem, smo združili v podatkovno zbirko, ki na ravni poteze povezuje fiziološke in šahovske značilke igralca. V sklopu petih eksperimentov smo modele strojnega učenja naučili in ovrednotili na obeh naborih značilk ter njuni kombinaciji. Za razlago napovedi smo uporabili metodo SHAP. Pri prepoznavanju igralca z nižjo oceno ELO so fiziološke značilke s točnostjo 62,4% presegle šahovske z 59,1%. Zgolj s fiziološkimi značilkami smo visoko prednost prepoznali s točnostjo 67,1%. Njihova vključitev v model s šahovskimi značilkami je točnost povečala z 72,5% na 75,5%. Pri napovedovanju napak smo s šahovskimi značilkami uspešno zaznali 54,8% napak s preciznostjo 26,9%, dodajanje fizioloških značilk pa napovedne uspešnosti ni izboljšalo. Modeli s fiziološkimi značilkami so pri vseh eksperimentih opazno presegli naključni klasifikator, pri napovedovanju napak pa je bil doprinos fizioloških značilk nekoliko manjši.

Language:Slovenian
Keywords:strojno učenje, napovedovanje šahovske uspešnosti, fiziološki signali, razložljivost modelov, personalizacija
Work type:Master's thesis/paper
Organization:FRI - Faculty of Computer and Information Science
Year:2026
PID:20.500.12556/RUL-191485 This link opens in a new window
Publication date in RUL:19.11.2026
Views:27
Downloads:2
Metadata:XML DC-XML DC-RDF
:
Copy citation
Share:Bookmark and Share

Secondary language

Language:English
Title:Predicting chess performance using physiological parameters
Abstract:
Stress is reflected in measurable physiological changes, which can affect decision making in mentally demanding activities. The goal was to determine whether physiological parameters can be used to predict chess performance and events, and whether they improve predictions based on chess features. Data collected at an official chess tournament was used to build a dataset combining each player's physiological and chess features for every move. In five experiments, machine learning models were trained and evaluated using physiological and chess features, and their combination. SHAP was used to explain predictions. When identifying the player with the lower ELO rating, physiological features achieved an accuracy of 62.4%, compared with 59.1% using chess features. Physiological features alone identified a substantial advantage with an accuracy of 67.1%. Adding them to a model using chess features increased accuracy from 72.5% to 75.5%. Using chess features, we detected 54.8% of mistakes with a precision of 26.9%, while adding physiological features did not improve predictive performance. Across all experiments, models using physiological features noticeably outperformed a random classifier, though to a lesser extent when predicting mistakes.

Keywords:machine learning, chess performance prediction, physiological signals, model explainability, personalization

Similar documents

Similar works from RUL:
Similar works from other Slovenian collections:

Back