Predictive analytics has become an essential area of research in health informatics. The availability of multi-source and multi-modal data in healthcare has made the disease prediction, diagnosis, and medication process more effective and reliable. However, the analysis and decision making have become challenging task, particularly when data is in multiple formats and from different sources. In this study, different frameworks have been proposed to handle multi-nature data at different levels for predictive analytics. Dimensionality reduction techniques have been applied to extract relevant features to enhance the analysis. To improve the performance of predictive analytics at different fusion levels, the potential benefits of multi-modal data have been discussed. Moreover, notable improvement in prediction accuracy has been observed through experimental evaluation of the proposed frameworks. Furthermore, the issues which have been found during dimension reduction and fusion approaches have also been highlighted.
High dimensional and multidimensional data is found in different applications for object classification, face image recognition, processing of electrocardiography (ECG) signals, medical images classification of disease, telecommunication, process and program monitoring. The preprocessing of high- and multidimensional data is required before applying analysis and machine learning algorithms. When dealing with multidimensional and 38high-dimensional data the following questions are generally raised: How can one reduce a data set having hundred or even thousands of features while preserving its core properties for analysis? How to visualise a data set having countless dimensions? How to eliminate redundant and noisy features? How to make precise classifications and accurate predictions in the presence of many redundant and missing features (Amaratunga and Cabrera 2016)?