housing-germany

Machine learning project for the Data Talks Club Machine Learning Zoomcamp 2023.

Problem Description

The real estate market in Germany is dynamic and influenced by various factors such as location, property type, size, amenities, and economic conditions. Accurately predicting rental prices is crucial for both property owners and potential tenants to make informed decisions. In this context, developing a reliable machine learning model to predict housing rent prices in Germany becomes paramount.

The objective of this project is to build a predictive model that can estimate the monthly rent for residential properties in Germany based on selected features.

Instructions

Dependency Management

pip install -r requirements.txt

Dataset

CorrieBar. (2019, October). Apartment rental offers in Germany, Version 6. Retrieved October 28, 2023 from https://www.kaggle.com/datasets/corrieaar/apartment-rental-offers-in-germany

To download the dataset follow the instructions on https://www.kaggle.com/docs/api

kaggle datasets download -d corrieaar/apartment-rental-offers-in-germany
unzip apartment-rental-offers-in-germany.zip

Explore Data (EDA) & Train Optimized Model with Notebook

Open the housing-eda.ipynb notebook.

Run Model Training Script

python scripts/train.py

Test Model Prediction Script

python scripts/predict.py

Model Deployment

Run App

python flask app/flask_app.py

streamlit run app/streamlit_app.py

Run Containerized App with Docker

docker build -t model_server .
docker run -p 5000:5000 -p 8501:8501 model_server

Streamlit Cloud Deployed

Visit the live app at housing-germany.streamlit.app