Data-driven predictive maintenance of distribution transformers

This dataset supports the results presented in the article titled \u201cData-driven predictive maintenance of distribution transformers: class imbalance and Bayesian optimization\u201d. It contains processed information on 16,000 distribution transformers located in the department of Cauca (Colombia), collected between 2019 and 2020 by the Compa\u00f1\u00eda Energ\u00e9tica de Occidente, the regional power distribution operator.The data used in this study corresponds to a transformed version of the dataset published by Bravo, Lozano, and \u00c1lvarez (2021) [https:\/\/doi.org\/10.1016\/j.dib.2021.107454], available under the Creative Commons CC BY 4.0 license. The applied transformations include data cleaning, variable encoding, and preparation for predictive modeling using machine learning techniques. This version is intended to facilitate the replication of the results described in the referenced article.