This paper utilizes the Support Vector Regression (SVR) and Artificial Neural Network (ANN) integrated with a GARCH model in analyzing volatility within the gold market. We used the root of mean square error to compare the performance between the econometric model and various ML-GARCH models in forecasting the stock price of COMEX Gold Futures. SVR model with RBF kernel is found to be the most successful model in predicting the future stock prices of COMEX Gold Futures with an extremely low RMSE value among the machine learning models. For the market risk evaluations, we found that the gold future market become less volatile during the COVID-19 pandemic as compared to before pandemic.