An interactive wpp application for predicting real estate prices across the United States, featuring a modern map interface, AI-powered chat, and economic factor inputs. This project is designed for demonstration.
- Interactive US Map: Explore states, counties, and zip codes with smooth navigation and 2D/3D views.
- Real Estate Price Prediction: Get price estimates for properties by zip code, using economic indicators such as gas prices, unemployment rate, interest rates, and inflation.
- AI Chatbot: Ask questions about any area and receive AI-generated insights.
- User Authentication: Secure login via Google (no manual sign-up required).
- Modern UI: Responsive, user-friendly interface with draggable popups and detailed overlays.
Can housing prices across U.S. regions be accurately predicted using economic indicators such as consumer price index (CPI), gas prices, interest rates, unemployment rates, mortgage rates, and other financial metrics within a machine learning framework?
Project Structure
• Unemployment Rates
• Housing Price Data
• Gas Prices
• Consumer Price Index (CPI)
• Mortgage Rate Data
• Interest Rates
We consolidated these datasets into a unified framework to facilitate a comprehensive analysis.
At the outset, our research was based on two primary datasets: housing prices and CPI data. Over time, additional datasets were incorporated to enrich the analysis. The datasets underwent thorough cleaning, including: • Removing missing values • Filtering data within the 2011–2022 timeframe • Excluding unnecessary attributes • Enhancing geographic precision with ZIP codes and latitude/longitude data
This structured approach ensured robust visualizations and meaningful regional insights.
Our project leveraged various programming tools, databases, and visualization platforms:
• Python (pandas, USzipcode, geopy, Nominatim API, pathlib/path)
• Jupyter Notebook
• CSV File Handling
• Python (NumPy, Matplotlib, Seaborn, Scikit-learn, TensorFlow, ... etc)
• Jupyter Notebook
git clone https://github.com/ananquay15x9/BDS.git
cd BDSThe backend API handles price predictions. Run the following command:
python api_merge.py- Make sure you have Python 3.x and the required dependencies installed.
- The API will start on
http://127.0.0.1:8000/by default.
- Open the file
index_login.htmlin your web 7AE0 browser. - Login using Google (recommended and required; there is no manual sign-up or database for user accounts).
- After logging in, you will be redirected to the main interactive map page.
-
Navigate the Map: Zoom, pan, and explore different regions of the US.
-
Search Locations:
Use the search box to find states, counties, or zip codes (restricted to US locations). -
Predict Prices:
- Click on the map or search for a location.
- Enter economic factors (or leave blank for default prediction).
- View predicted real estate prices for the selected zip code.
-
Chat with AI:
- Use the AI chatbot to ask about the advantages or characteristics of any area.
- Data Coverage:
Not all US zip codes are available for prediction due to limited data. If a zip code is not found, an error message will be shown. - Authentication:
Only Google login is supported for simplicity and security. - For Local Use:
This app is intended for local/demo use. For production, further security and deployment steps are required.
- Frontend: JavaScript, MapLibre GL JS, HTML, CSS
- Backend: Python (FastAPI or Flask, depending on your
api_merge.py) - Geocoding: Stadia Maps API (US-only search)
- Authentication: Google OAuth
For questions or feedback, please open an issue or contact me via GitHub.