Performance Evaluation of Whale Optimization Algorithm and Bald Eagle Search in Predictive Housing Price Modeling

Abstract

Accurate housing price prediction remains a critical challenge in real estate analytics, requiring models capable of capturing nonlinear relationships among structural, amenity, and economic features. This study evaluates the performance of the Whale Optimization Algorithm (WOA) and the Bald Eagle Search (BES) when integrated into a K-Nearest Neighbors (KNN) regression framework. Using a Kaggle housing dataset enriched with attributes such as area, bedrooms, bathrooms, and local economic indicators, the models were optimized and benchmarked against a baseline KNN. Experiments conducted in Google Colab assessed predictive accuracy using Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and the coefficient of determination (R²). Results show that the baseline KNN achieved an MAE of 193.97, RMSE of 265.23, and R² of 0.5705, reflecting moderate accuracy. The WOA-KNN model slightly increased MAE to 196.33 but reduced RMSE to 261.71, improving R² to 0.5818. Similarly, BES-KNN produced identical values to WOA-KNN, confirming comparable optimization outcomes. These findings demonstrate that metaheuristic-driven models enhance predictive stability and explanatory power, with WOA-KNN showing the most consistent error distribution. The study contributes to bridging the gap in comparative evaluations of WOA and BES, offering practical insights into the trade-off between exploration and exploitation strategies for reliable housing price forecasting.