This study addresses the challenges of stock market forecasting by introducing a hybrid model, RBM-MLP, combining deep learning (DL) and neural network techniques. The goal is to predict closing prices for Alphabet Inc. (Google) and Meta Platform (Facebook) stocks, acknowledging the complexity of the time series problem due to inherent uncertainty and various influencing factors in stock data. Traditional time series methods face limitations in achieving accurate predictions, prompting the adoption of DL and machine learning models. The proposed hybrid model integrates a DL technique, Restricted Boltzmann Machine (RBM), and a neural network, Multi-Layer Perceptron (MLP), for enhanced forecasting accuracy. The RBM-MLP model is evaluated using stock data from Alphabet Inc. and Meta Platform. Dataset preprocessing tackles standardization and normalization challenges, while technical indicators are applied to extract features. RBM is then employed for further feature extraction. The dataset is strategically partitioned for training and testing the MLP model in the prediction analysis. Evaluation metrics such as Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and the correlation coefficient R2 are utilized to assess the performance of the RBM-MLP model. These metrics provide valuable insights into the model’s accuracy in predicting stock prices, contributing to the advancement of stock market forecasting methodologies. In summary, this research presents and evaluates the effectiveness of the RBM-MLP hybrid model, showcasing its potential to improve prediction accuracy in stock market forecasting.