Work

Olympic Medal Prediction Based on Few Sample Learning

Year: 2025

Type: conference-paper

Abstract: We introduce a novel hybrid GAN model (E-Deep SVDD-Synthetic data generator) that forecasts the path-breaking nations who have won little in the past Olympic Games. The new idea of generating more samples using synthetic generation has overcome the limitation of having few historical examples, computed possible countries' success while identifying possible success according to historical records, and gave an excellent prediction of upcoming 2028 Olympic medalists like Angola (82.55%), Eritrea (54.17%), and Honduras (51.55%). The validation was done using the loss curve and accuracy curve, which indicates the reliability.

Source: International Conference on Applied Mathematics, Modelling and Intelligent Computing (CAMMIC)

DOI: 10.1145/3745533.3745575

Authors: Sijiu Yu, Xiaohua Wang, Yuxuan Xiao

Institution: Chongqing University

Language: English

Topic: Sports Analytics and Performance

Subfield: Economics and Econometrics

Field: Economics, Econometrics and Finance

Domain: Social Sciences

Open Access status: hybrid

OpenAlex ID: W4413602185

FWCI:

Cites: 1

Cited by:

Related to: 10