Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Predicting House Rental Prices in Ghana Using Machine Learning

Philip Adzanoukpe

TL;DR

The paper tackles the opacity of Ghana's rental market by building a data-driven approach to predict house rental prices from Tonaton listings, comparing Linear Regression, Random Forest, SVR, XGBoost, and CatBoost. It leverages feature engineering (geolocation and amenities) and a robust evaluation framework using $R^2$, MSE, MAE, and RMSE, with CatBoost achieving the best performance at $R^2 \approx 0.876$. Key insights show location-based features and property attributes (bedrooms, bathrooms, furnishing) as primary price drivers, validated through residual analyses and feature-importance assessments. The work offers a practical tool for tenants, landlords, and policymakers to understand and predict rents, and points to future enhancements including temporal data and regional variation to improve generalizability.

Abstract

This study investigates the efficacy of machine learning models for predicting house rental prices in Ghana, addressing the need for accurate and accessible housing market information. Utilising a comprehensive dataset of rental listings, we trained and evaluated various models, including CatBoost, XGBoost, and Random Forest. CatBoost emerged as the best-performing model, achieving an $R^2$ of 0.876, demonstrating its ability to effectively capture complex relationships within the housing market. Feature importance analysis revealed that location-based features, number of bedrooms, bathrooms, and furnishing status are key drivers of rental prices. Our findings provide valuable insights for stakeholders, including real estate professionals, investors, and policymakers, while also highlighting opportunities for future research, such as incorporating temporal data and exploring regional variations.

Predicting House Rental Prices in Ghana Using Machine Learning

TL;DR

The paper tackles the opacity of Ghana's rental market by building a data-driven approach to predict house rental prices from Tonaton listings, comparing Linear Regression, Random Forest, SVR, XGBoost, and CatBoost. It leverages feature engineering (geolocation and amenities) and a robust evaluation framework using , MSE, MAE, and RMSE, with CatBoost achieving the best performance at . Key insights show location-based features and property attributes (bedrooms, bathrooms, furnishing) as primary price drivers, validated through residual analyses and feature-importance assessments. The work offers a practical tool for tenants, landlords, and policymakers to understand and predict rents, and points to future enhancements including temporal data and regional variation to improve generalizability.

Abstract

This study investigates the efficacy of machine learning models for predicting house rental prices in Ghana, addressing the need for accurate and accessible housing market information. Utilising a comprehensive dataset of rental listings, we trained and evaluated various models, including CatBoost, XGBoost, and Random Forest. CatBoost emerged as the best-performing model, achieving an of 0.876, demonstrating its ability to effectively capture complex relationships within the housing market. Feature importance analysis revealed that location-based features, number of bedrooms, bathrooms, and furnishing status are key drivers of rental prices. Our findings provide valuable insights for stakeholders, including real estate professionals, investors, and policymakers, while also highlighting opportunities for future research, such as incorporating temporal data and exploring regional variations.
Paper Structure (24 sections, 9 equations, 8 figures, 2 tables)

This paper contains 24 sections, 9 equations, 8 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Literature Review
  3. Housing Market in Ghana
  4. Machine Learning for House Price Prediction
  5. Bridging the Gap
  6. Methodology
  7. Linear Regression
  8. Random Forest
  9. Support Vector Machines
  10. Extreme Gradient Boosting (XGBoost)
  11. CatBoost
  12. Implementation
  13. Model Evaluation
  14. R-squared ()
  15. Mean Squared Error (MSE
  16. ...and 9 more sections

Figures (8)

  • Figure 1: Distribution of rental prices in Ghana
  • Figure 2: Distribution of rental listings in different locations in Ghana
  • Figure 3: Illustration of rental prices in different locations in Ghana
  • Figure 4: Distribution of top amenities provided in rental houses
  • Figure 5: Scatter plot of actual vs predicted rental prices
  • ...and 3 more figures