In the financial markets, volatility forecasting is a crucial area of study. Much work has been devoted to developing volatility models, as more accurate forecasts enable more advantageous option pricing and risk management. Various approaches (e.g., econometrics and machine learning) exist to predict stock market returns. However, it’s not easy to find the approach that works best. In this paper, we perform an in-depth analysis of the predictive accuracy of several machine learning algorithms (SVR, Gradient Boosting, LSTM, and GRU) to predict the volatility of the Dow Jones Industrial Average Index. Taking into account data-spying bias, four different measures are applied to examine the predictive ability of each model. The study aims to compare machine learning algorithms for volatility predictions. To assess the efficacy of several ML models in predicting DJIA volatility. To provide information about the merits and disadvantages of each strategy.