Abstract—This paper presents a novel approach to spatial arbitrage in cryptocurrency markets by integrating machine learning (ML) models with Explainable Artificial Intelligence (XAI) techniques. We establish a robust framework for estimating price differences between exchanges using historical returns, technical indicators, and unique exchange data. Our results demonstrate increased profitability and reduced risk for arbitrage strategies, particularly in less liquid cryptocurrency pairs. We propose a scalable, interpretable feature selection mechanism to facilitate dynamic arbitrage decisions in decentralized cryptocurrency markets.
|
DOI:
10.17148/IJARCCE.2024.131239
