Utilizing a variety of machine learning techniques to improve accuracy and comprehension, this study investigates the challenging topic of housing price prediction. The research covers Dhaka and Chittagong, two significant real estate hotspots, over a 3-year period using a selected dataset from http://bdproperty.com/ . The primary prediction characteristics in the dataset are location, number of rooms and bathrooms, area size, property type, and timestamp. To provide a solid foundation for machine learning efforts, the research begins with comprehensive pre-processing and data analysis. R-squared (R2) and Mean Squared Error (MSE) are used to evaluate a range of models, such as XGBoost, Random Forest, and Linear Regression. The results highlight the significance of feature scaling and other preprocessing techniques while showcasing Random Forest’s outstanding performance. The research highlights how machine learning has the ability to provide useful insights to stakeholders in the housing market, with consequences having an impact on real-world applications. The study concludes with important discoveries and suggests directions for future research, like adding more variables, doing temporal analysis, using advanced ensemble methods, exploring new areas geographically, and enhancing the interpretability of the model. This work essentially adds to the changing field of real estate prediction by demonstrating the revolutionary potential of machine learning for comprehending and forecasting housing dynamics.