Missing data imputation with Harmony Search Algorithm

• Published in: International Journal of Combinatorial Optimization Problems and Informatics Vol. 16; no. 4; pp. 381 - 398
• Main Authors: Oviedo Salas, Edgar Alberto, Balderas-Jaramillo, Fausto Antonio
• Format: Journal Article
• Language: English
• Published: Jiutepec International Journal of Combinatorial Optimization Problems & Informatics 12.10.2025
• Subjects: Data science; Machine learning; Missing data; Prediction models; Root-mean-square errors; Search algorithms; Water resources
• ISSN: 2007-1558; 2007-1558
• DOI: 10.61467/2007.1558.2025.v16i4.999

Incomplete data poses a significant obstacle in data science and machine learning, influencing model outcomes and occurring commonly across domains such as health, nutrition, electricity, agriculture, chemistry and water resources. Missing data refers to the absence of information for one or more variables in a dataset. Accurate imputation is therefore crucial to ensure the reliability and validity of analyses and predictive models. This study proposes a Harmony Search Algorithm (HSA) to address missing-data imputation, emphasising its flexibility and adaptability. The approach seeks the best imputations by minimising error metrics such as MAE, MSE and RMSE. Computational tests indicate that HSA is a promising method for imputing missing data in a range of contexts.