DOI number:

10.1007/s00500-023-08841-2

Affiliation of Author(s):

(1) School of Computer Science, Wuhan University, Wuhan; 430072, China

Journal:

Soft Computing

Abstract:

The healthcare data is generated through blockchain databases, web-based databases and mobile applications-based data. This research is using machine intelligence to draw inferences from the healthcare data to generate futuristic data from it. In other words, the prediction of future health issues can be made using machine intelligence based on the prevailing trends of historical data. Concept drift refers to the phenomenon of dynamic changes in data distribution, which is commonly present in mobile health monitoring data that are susceptible to environmental influences, hindering medical analysis and decision-making. The research work proposes a latent distribution self-evolving architecture (LDSA) based on neuronal variants and latent distribution recombination (LDR) to achieve data representation and prediction in the context of concept drift of futuristic healthcare data. The genetic mechanism of LDSA can adaptively handle concept drift in data and drive latent data distribution towards the future, eventually achieving prediction of future data. The machine learning-based intelligent mechanism can predict the future cases of diabetic patients in the world by analysing the current and historical data of diabetic patients in the world. We also design an evolution loss function to optimize the model performance and improve the pattern generalization ability. Experimental results show that LDSA achieves state-of-the-art performance on mobile health datasets and has better drift adaptation than popular baseline methods. The proposed machine intelligence-based algorithm is able to predict the future data by learning from the historical data features. The healthcare industry will be greatly benefitted from this research work to predict the futuristic ailments in the population on the basis of changing healthcare parameters of the current population. © 2023, The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.

Co-author:

Fang(1),Yu, Wei(1), Yu, Li, Shijun(1)

Translation or Not:

no

Date of Publication:

2023-01-01