Skip to content

House Price Predictor

A machine learning system that estimates residential property sale prices from structural and quality features. Built and trained on the Ames Housing dataset, a standard benchmark for regression modelling with 79 explanatory variables describing almost every aspect of residential homes in Ames, Iowa.

What It Does

Given a set of property characteristics — floor area, build year, quality rating, garage size, and others — the system returns a predicted sale price. A Streamlit web interface wraps the model so non-technical users can interact with it directly without touching any code.

Three regression models were trained and evaluated. The Random Forest model was selected for production based on its performance on the held-out test set.

Source Repository

The source code, trained model files, and original notebook live in the archived House-Price-Prediction repository. This documentation site is maintained separately as a standalone docs project.

Section What it covers
Getting Started How to run the app locally
Methodology Dataset, feature selection, model training
App Reference Input fields, expected values, output
Results Model comparison and selection rationale

Stack

  • Modelling — scikit-learn (Linear Regression, Ridge, Random Forest)
  • Interface — Streamlit
  • Docs — MkDocs + Material theme, deployed via GitHub Actions