High-dimensional machine learning to predict hospital readmission among older people with chronic kidney disease: abridged secondary publication.

1. A novel factorisation-based machine-learning model, which is Hong Kong-weighted with fuzzy partition, is proposed to predict the risk of hospital readmission among patients with chronic kidney disease. 2. This is one of the first machine-learning models for readmission prediction based on territorywide data in Hong Kong. 3. Application of this model may help reduce the cost of hospital management and improve the quality of life among patients with chronic kidney disease.