OPTIMIZED TRANSFER LEARNING BASED DEMENTIA PREDICTION SYSTEM

According to the World Health Organization (WHO), dementia is “an umbrella term for several diseases affecting memory, and behaviour that interferes with a person’s ability to maintain their daily living activities. It is not a normal part of aging”. Dementia is caused by physical changes in the brain. Drugs are available to alleviate some of the symptoms, but they do not cure them. In the existing system, it is difficult to identify if a person is suffering from Dementia. It can be only done with the help of clinical history and by knowing if the person has some genetic disorder. Many machine learning algorithms like SVM, KNN were used to overcome the “quantifiability of the stages in dementia” but there was a decline in the accuracy. To overcome the limitations of the existing system, the proposed system was structured in a way to use classification based on the features like Years of Education, Socio- Economic Status, Mini-Mental State Examination, Clinical Dementia Rating, Estimated Total Intracranial Volume, Whole Brain Volume, Atlas Scaling Factor. The purpose of this study was to provide a new clinical tool based on ensemble learning techniques like Random Forest, AdaBoost and Light GBM which can increase the accuracy of the final outcome.