Designing A Cloud-Based Framework using Data Mining for Healthcare Services in Remote Areas

Advancement is recently made inthe medicinal field producesnew innovative technologies to the healthcare sector and medical services. Access to quality healthcare is a major problem in remote areas, with a doctor-to-patient ratio as high as 1:20,000 which is far above the recommended ratio of the World Health Organization (WHO). This has been antagonized by a lack of access to critical infrastructures such as the health care facilities, roads, electricity, and many others. To watch basic medical parameters for identifying the abnormalities within the first stage of chronic diseases need regular interval hospital visits, which can be a comparatively costly and time-consuming process. Rare availabilities of doctors or medical centers, ignorance of the people, and proper care at the right time are the prime causes of great medical concern, which leads to unexpected death. This work is an attempt to solve basic health problems and take advice from registered medical experts for the betterment of the targeted community.Rapid development in the cloud environment, health care services are reasonable to the people in remote areas. It is necessary to predict the disease and connect with the doctor to get an early diagnosis of disease. The imperative goal of the paper is to develop a cloud-based framework using data mining to enhance healthcare in remote areas. The cloud-based framework is designed and simulated by using Matlab R2018b. Fast SearchGrowing Self Classifier (FS-GS) data mining classifier is developed to separate the data from the cluster to correlate the symptoms of patients with specialists. The classifier parameters like accuracy, precision, recall, sensitivity, and specificity are analyzed to compare the efficiency of the proposed algorithm with the other data mining algorithms like Naive Bayes, Random Forest, K Nearest Neighbor, and Support Vector Machine Linear. The proposed FS-GS data mining Classifier obtains an accuracy of 92%, the precision of 90.01%, recall of 90.06%, the sensitivity of 94.91%, and specificity of 92.6%. For the effectiveness, the proposed algorithm is compared with the various mining data classification algorithms to showmance of the proposed algorithm. Ultimately, the result shows the proposed algorithm scores higher outputs than all other algorithms in real time scenarios respectively.