MSc.Thesis Defense:Atra Zeynep Bahçeci

BUSINESS POINT OF INTEREST RECOMMENDATION WITH REINFORCEMENT LEARNING

Atra Zeynep Bahçeci
Data Science, MSc. Thesis, 2024

Thesis Jury

Prof.Selim Saffet Balcısoy (Thesis Advisor),

Asst. Prof. Onur Varol,

Asst. Prof. Mehmet Ali Ergün

Date & Time: 17th of December, 2024 –  11.00 AM

Place: FASS 1001

Keywords : business location selection, reinforcement learning, deep q-learning,

location intelligence

Abstract

The importance of location for business success cannot be overstated. Existing approaches to the business location selection problem often involve creating extensively tuned models specific to the geographical and economic climate being analyzed, and thus suffer from limited generalization across diverse scenarios. This thesis proposes a novel Deep Q-Learning framework for business location recommendation that can be trained in one geographic area and applied to another without requiring further training or tuning. Comprehensive experiments on real-world data demonstrate the superior generalizability of the proposed recommendation framework, outperforming the well-established Huff gravity model by 15.33% in profits, with an average profit realization of 78.02% compared to the best-case scenario. Empirical results indicate that variation in training data must be as high as the variation in the test data for the framework to be successfully applied to other locations despite discrepancies between the characteristics of the cities. The proposed approach offers a highly generalizable and easily applicable solution to the business location selection problem, providing a strong alternative to gravity-based models.