puja-urmi / house-price-prediction-sklearn Goto Github PK

Welcome to our repository dedicated to the prediction of residential property prices. We have meticulously implemented and rigorously tested a diverse range of machine learning regression models, including Linear, Lasso, Ridge, Decision Tree, K Nearest Neighbor (KNN), Support Vector Regression (SVR), and Ensemble techniques like Gradient Boosting and Random Forest. The dataset used is available in the "Data" folder. Our analysis centers on employing these models on a unified dataset to predict property prices based on crucial features such as size, room count, and more.

In the notebook script titled "main.ipynb," the project is presented along with its detailed findings, and the final section includes a comparison of these findings. This comprehensive exploration aims to fine-tune predictive accuracy, specifically for price estimation. In the "code" folder, you will find the implemented codes to process the data and a Python file of 'main.ipynb' saved as 'regressors.py'. Through exhaustive evaluations and careful parameter tuning using grid search methodologies, we've aimed to identify the best parameter sets tailored to our dataset's characteristics. The techniques applied to analyze house features could be valuable for similar data types, providing insights.

Our findings for this dataset highlight the exceptional effectiveness of tree-based models and ensemble techniques when compared to other methods. Dive into our repository to explore our journey and extract insights into the most efficient strategies for predicting house prices.