When research fields such as biological and medical arenas are considered, computer science-based applications have been utilized on an increasing basis. In the process, data has accumulated in significant amounts. In particular, most of the big data that has resulted has accrued from experimental and medical results that are reported. The eventuality is that in the healthcare platforms, data mining techniques have been employed towards timely predictions of conditions with which patients present, a process achieved through data analysis. Also, the data analyses have strived to discern correlations among disease-related variables or parameters in otherwise voluminous information. It is also notable that data mining has received increasing adoption in the healthcare field because of the perceived ability to provide room for informed decision-making, reliable treatment decisions, and the detection of potential fraudulence when it comes to health care service payments. While these interplays point to a promising trend regarding the utilization of data mining and disease severity prediction, the eventual bid data has proved to be so complex that the use of traditional methods of data processing and analyzing is deemed less effective. In addition, given data warehouse, the use of traditional methods towards data extraction makes it difficult to discern some of the hidden intersections, pointing to the importance of a new approach for data classification and disease severity prediction. The aim of this review paper was to demonstrate how the use of classifiers in the form of machine learning algorithms could be utilized in the prediction of the severity of diseases. Indeed, the context of the research entailed warehouse environments. To have an ideal kernel established towards data classification (that would then enable additional diagnostic tests), some of the machine learning algorithms that were investigated included SVM (support vector machine) and ANN (artificial neural network). From the review outcomes, it was established that when machine learning algorithms are used to obtain trend similarities in disease severity through predictive analytics, similar severity magnitudes of the given diseases could be projected in terms of their probabilities of and trends in occurrence. Also, the review outcomes demonstrated that disease severity prediction through machine learning algorithms is an insightful practice because it allows for the prediction of patients’ conditions, as well as the prevision of room for optimal decision-making regarding treatment modalities to be adopted by healthcare firms.