The Black-Scholes (BS) model has been used as the foundation for pricing options in the stock market for years. However, this mathematical model is very simple, which limits its capability to predict the present market dynamics for stock prices. In this work, stock price data of twenty years of a leading energy company is used to predict option prices directly from the BS closed-form solutions. Afterwards, four different machine learning (ML) models, namely linear regression, XGBoost regression, support vector machine, and random forest regression, are used to forecast the stock prices 3 days and 10 days ahead of option maturity. The performances of all these ML models are compared using the coefficient of determination R², root mean square error (RMSE) and mean absolute error (MAE). Among all, the XGBoost algorithm achieved the highest R² value of 0.85. Since this ML model showed the closest fit to the market stock prices, the residuals of option pricing using XGBoost have been calculated and hence compared with BS as a justification of the estimation. The result of the residue is around ± 0.11. It indicates superior predictive capability compared to other approaches. The lower RMSE and MAE values further confirm that XGBoost provides the most accurate fit between predicted and actual option prices by analyzing the stock prices of the company. It is anticipated that this study will be helpful in preventing sudden stock market crashes and improving the stability of the market.