ℹ️ About This Project
The Bangalore House Price Predictor estimates residential property prices in Bangalore using a Machine Learning model trained on real housing data. It's built to help buyers, sellers, and researchers get a quick, data-driven sense of what a property might be worth.
🧪 Methodology
-
1
Data Cleaning
Removed missing values, inconsistent size formats, and irrelevant columns from the raw dataset.
-
2
Outlier Removal
Filtered out properties with implausible price-per-sqft values and bathroom-to-bedroom ratios.
-
3
One-Hot Encoding
Converted the location column into one-hot encoded features so the model can use categorical location data.
-
4
Train/Test Split
Split the cleaned dataset into training and testing sets to evaluate the model on unseen data.
-
5
Linear Regression
Trained a Linear Regression model on area, location, BHK, and bathroom count to predict price.
-
6
Evaluation
Measured model accuracy using cross-validation to confirm consistent performance across folds.
-
7
Served via Flask API
Wrapped the trained model in a Flask API so predictions can be served instantly to this web app.
📊 Dataset at a Glance
🛠 Tech Stack
📌 Model Notes
Predictions from this tool are statistical estimates based on historical data patterns, not exact appraisals. Actual market prices can vary due to factors the model doesn't capture, such as property condition, exact floor, view, amenities, and current market sentiment. Use these estimates as a starting point for research, not as a substitute for a professional valuation.