In the dynamic landscape of modern markets, where volatility reigns supreme, the pair trading strategy shines as a beacon of stability and profitability. By capitalizing on correlated movements between carefully selected asset pairs, investors navigate uncertainty with precision, maximizing returns and mitigating risks. Yet, the paramount challenge lies in identifying the most suitable pairs for pair trading. This project aims to identify such pairs through the utilization of machine learning methods. Employing a multi-step approach, the research begins with foundational statistical analyses such as correlation and cointegration to identify initial pairs. Subsequently, advanced machine learning techniques including principal component analysis (PCA), K-means clustering, density-based spatial clustering of applications with noise (DBSCAN), and hierarchical clustering are employed to further refine pair selection. These methodologies enable the classification of assets into meaningful clusters based on their price behaviors, facilitating more informed trading decisions. Through empirical analysis, distinct clustering patterns emerge: K-means clustering yields 12 clusters, DBSCAN identifies 6 clusters, and hierarchical clustering produces 5 clusters. Leveraging these clusters, pair trading strategies are constructed and evaluated against historical market data to assess their performance.