Adaptive Swarm Balancing Algorithms for rare-event prediction in imbalanced healthcare data

• Published in: PloS one Vol. 12; no. 7; p. e0180830
• Main Authors: Li, Jinyan, Liu, Lian-sheng, Fong, Simon, Wong, Raymond K., Mohammed, Sabah, Fiaidhi, Jinan, Sung, Yunsick, Wong, Kelvin K. L.
• Format: Journal Article
• Language: English
• Published: United States Public Library of Science 28.07.2017; Public Library of Science (PLoS)
• Subjects: Adaptive algorithms; Algorithms; Analysis; Biology and Life Sciences; Cancer; Classification; Computer and Information Sciences; Computer engineering; Computer science; Data analysis; Data mining; Data processing; Datasets; Delivery of Health Care; Disease; Disease control; Experiments; Health care; Health care information services; Heuristic methods; Information science; Medical diagnosis; Medical research; Medicine and Health Sciences; Methods; Optimization; Particle swarm optimization; Physical Sciences; Research and Analysis Methods; Run time (computers); Sampling techniques; Swarm intelligence; Australia; Canada; Macao
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0180830

Clinical data analysis and forecasting have made substantial contributions to disease control, prevention and detection. However, such data usually suffer from highly imbalanced samples in class distributions. In this paper, we aim to formulate effective methods to rebalance binary imbalanced dataset, where the positive samples take up only the minority. We investigate two different meta-heuristic algorithms, particle swarm optimization and bat algorithm, and apply them to empower the effects of synthetic minority over-sampling technique (SMOTE) for pre-processing the datasets. One approach is to process the full dataset as a whole. The other is to split up the dataset and adaptively process it one segment at a time. The experimental results reported in this paper reveal that the performance improvements obtained by the former methods are not scalable to larger data scales. The latter methods, which we call Adaptive Swarm Balancing Algorithms, lead to significant efficiency and effectiveness improvements on large datasets while the first method is invalid. We also find it more consistent with the practice of the typical large imbalanced medical datasets. We further use the meta-heuristic algorithms to optimize two key parameters of SMOTE. The proposed methods lead to more credible performances of the classifier, and shortening the run time compared to brute-force method.