Abstract:
"Accurately predicting house selling prices is crucial for both buyers and sellers in the real estate
market today. However, existing methods often struggle to consider diverse factors and adapt to
dynamic market conditions. This research proposes a novel house selling price prediction system
based on machine learning using linear regression models. Author has created a own dataset of
real estate especially for Sri Lankan real estate market context transactions, incorporating various
features such as location, property characteristics, and market trends.By using advanced machine
learning techniques, the author developed a model that achieves significantly higher accuracy
compared to traditional approaches. These findings demonstrate the potential of machine
learning to improve the efficiency and transparency of the real estate market by providing more
reliable price estimations. Further research directions are explored to enhance the system's
capabilities and expand its applicability to different market contexts.
The system gets regression models to forecast house prices accurately. Initial results of the
prototype indicate 80% accuracy for the regression model. Additionally, by performing
GridSearchCV pointed further enhancements, with the linear regression model achieving an
accuracy of 82.36%, for decision tree model got 82.60% and 82.36% for lasso model. The study
discusses the methodology used for data collection, preprocessing, model training, and
evaluation, highlighting key factors which are influencing house prices and model performance
metrics.
In conclusion, the promising results obtained from the linear regression model indicate its
effectiveness in accurately forecasting house prices. Further research directions are explored to
enhance the system's capabilities and expand its applicability to different market contexts."
