In recent years, the analysis and forecasting of real estate prices and transaction patterns in metropolitan regions have gained significant attention due to the dynamic nature of urban development. Traditional methods such as statistical modeling and survey-based approaches often fall short when processing large-scale, high-dimensional datasets. This project presents a deep learning-based framework for analyzing real estate market trends using publicly available government data. The dataset includes features such as property type, estimated value, sale price, locality, and transaction details. After preprocessing and standardization, machine learning and deep learning models such as Random Forest, Multi-Layer Perceptron (MLP), XGBoost, CatBoost, and Gradient Boosting were implemented to predict property prices. The study emphasizes the model's adaptability to evolving data patterns and its superior performance compared to conventional techniques. Evaluation metrics such as R², MAE, MSE, and RMSE were used to validate the predictive accuracy of the models. The results demonstrate the potential of integrating deep learning techniques into the real estate domain to support more accurate and efficient decision-making.

Keywords: Real Estate Price Prediction, Deep Learning, Machine Learning, XGBoost, CatBoost, Gradient Boosting, Property Transactions, Urban Housing Market, Price Forecasting, Data Standardization, Regression Models, MAE, RMSE, R².