Application of Machine Learning in Investment Risk Assessment and Decision Optimization

In a time when markets are volatile and data complex, financial decisions made are subject to investment risk assessment, which is one of the most important steps. This study aims towards exploring the use of advanced machine learning models like Random Forest, Gradient Boosting and Support Vector Machines (SVM), in determining investment risks and for decision-making purposes. A quantitative research methodology with simulation-based data generated via Monte Carlo simulation techniques is used, and models are evaluated, based on various metrics like accuracy, F1-score, AUC-ROC, MSE, RMSE, and MAE. It was found that Random Forest was superior to the other models, obtaining an accuracy of 0.9367, an F1-Score of 0.9385 and an AUC-ROC of 0.9794. Moreover, the Random Forest has also minimum MSE (0.063), RMSE (0.251) and MAE (0.063) among all models. The comparative performance also further depicts Random Forest’s superiority in predicting investment risk. Random Forest or SVM should be integrated with the financial institution’s decision-making process to enhance those risks. This research has practical implications in that it enables improving investment strategies with better machine learning models, however, simulation-based data come at a limitation that may be unable to reflect all the real-world market complexities.