An Intelligent Metaheuristic Technique Assisted Feed Forward Neural Network (FFNN) for Alzheimer Disease Classification System

One of the prevalent and general kind of dementia is Alzheimer Disease (AD). Only few people detects the incidence of disease at the early stage, since there is no specific symptom. Distinct reason for cause and cure of AD is not identified. When the disease is identified at the early stage, it can be managed and treating the patient also effective. This context necessitated early identification of disease with the assistance of computational technologies. The performance of disease classification is influenced by the dimensionality, irrelevant feature selection and misclassification. In this research work, shortcomings in the existing disease identification is improved by introducing optimization and deep learning techniques. The issues in handling dimensionality is handled by Linear Discriminant Analysis (LDA) and optimal features are retrieved by Fireworks Algorithm (FA). The optimal feature are trained and disease is classified using Feed Forward Neural Network (FFNN). The classification result is investigated by comparing with the existing state-of-art classification algorithms. The proposed LDA-FA-FFNN outperforms the existing approaches and achieves 93% accuracy