Classification of Money Market Mutual Fund Products in Indonesia By Using Mahalanobis Distance and Manhattan Distance

This study aims to classify money market fund products listed and supervised by the Financial Services Authority (OJK) with minimal classification error. Mahalanobis distance and Manhattan distance were employed to classify these products. Data was sourced from the Indo Premier Online Technology (IPOT) application. Variables utilized in this research include percentage return, Sharpe ratio, unit growth, and Asset Under Management (AUM) . Additionally, Principal Component Analysis (PCA) was employed to reduce data dimensionality by linearly combining correlated original variables into new variables (principal components). PCA was used to visualize data with more than three dimensions. Based on the principal component analysis, the first two principal components captured 74.43% of the original data information, while the first three principal components captured 98.94%. Classification results using three principal components and standardized data showed the same error rates: 13.33% for Mahalanobis distance and 6.67% for Manhattan distance. For the two principal components, both Mahalanobis and Manhattan distances resulted in an error rate of 13.33%. Therefore, Manhattan distance is the most effective method for classification. Forecasting results indicate that mutual fund A is a good investment choice, while mutual fund B is a poor one. Keywords: Mahalanobis distance; Manhattan distance; Principal Component Analysis