Diversified Crypto Assets Portfolio Optimization Using K-Means Clustering Algorithm And The Efficient Frontier
Rizki Setiawan, Muhammad Saiful Hakim · 2023
The objective of this study is to examine the possibility of utilizing the machine learning approach (K-means clustering) for finding efficient frontier based on the risk and return association. Using 84 crypto coins included in CMC Crypto 200, this research applies k-means clustering to group the research data. The result show that crypto coins are grouped into 6 clusters. Eight crypto coins selected to construct portfolio investment based on Sharpe ratio and market capitalization for portfolio optimization stage. Our testing model shown that all of our portfolio model outperforms the market especially compared with CMC Crypto 200 Index performance.