Incorporating game theory with soft sets for better decision making

Solving uncertainty is a challenge for decision-making. Soft set theory aims to aid complex decision-making when multiple uncertainty variables are involved. To solve classification problems with the existence of uncertainty, we adopted three-way classification instead of binary classification. It introduces a boundary region to handle scenarios in which a number of objects cannot be categorized as either positive or negative with a high degree of certainty. The three-way classification problem involves multiple experts. Each expert may produce a different three-way classification outcome based on their available information and expertise. We introduced a gametheoretic soft set model to address the fusion of partial information which is available to certain experts and resolve conflicts among experts when determining the final three-way classification. It uses a soft set to represent experts and formulates a game among parameters of the soft set. The model is utilized to establish measurement thresholds for parameters. The experiment shows the model is capable of striking a balance among different parameters, resulting in a decrease in misclassification error in an environment involving uncertainty. Furthermore, the extent of the decrease can be fine-tuned by adjusting the ratio between the cost for misclassification error and the cost for undecided error. Based on the user’s specified target misclassification error and undecided error, our model can help determine an appropriate ratio.

Name thesis:degree_name

Master of Science (MSc)

Level thesis:degree_level

Master's

Discipline thesis:degree_discipline

Computer Science

Grantor dc:publisher

Faculty of Graduate Studies and Research, University of Regina

Year dc:date.issued

2024

Author dc:creator

• Li, Chenqi

Advisor dc:contributor.advisor

• Yao, JingTao

Committee member dc:contributor.committeemember

• Hepting, Daryl

Language dc:language.iso

en

DOI

https://doi.org/10.82465/3816

OAI identifier oai:identifier

oai:uregina.scholaris.ca:10294/16473