Abstract:
Alzheimer’s disease (AD) is a cognitive illness that commonly occurs in 65 years old or above people, which destroys the neurons and many parts of the brain. This disease has no cure, treatment can only slow the damage progression and finally, death occurs. So the early detection of this disease is essential. AD affects 40 million people worldwide. The number of AD patients is constantly increasing. So, it is necessary to identify the progression of AD. There are no single test has been developed to diagnose this disease. Different clinical methods-Mini-Mental State Examination (MMSE), Montreal Cognitive Assessment (MoCA), Alzheimer’s disease Assessment Scale (ADAS), and neuroimaging techniques- positron emission tomography (PET), Magnetic Resonance Imaging (MRI), diffusion tensor imaging (DTI) is used to detect this disease. In this paper, Synthetic Minority Oversampling Technique (SMOTE) is used to oversample the dataset. And different machine learning algorithm such as support vector machine (SVM), K-Nearest Neighbor (K-NN), and Naïve Bayes (NB) to detect the stage of AD based on the different clinical data, cognitive data, brain data etc. 40 features were selected for training the model. Dataset obtained from Alzheimer’s disease Neuroimaging Initiative (ADNI) database. The accuracy of the models are high. I obtain 85%, 93%, and 96% accuracy from k-nearest neighbor, support vector machine, and naïve bayes algorithm. Naïve Bayes provide highest accuracy.