Alzheimer Disease Prediction Using Recursive Feature Elimination and Artificial Neural Network

Alzheimer's Disease (AD) is a devastating neurodegenerative disorder that affects a significant portion of the aging population worldwide. Early prediction and detection of Alzheimer's disease are crucial for providing timely interventions, personalized treatment plans, and improved patient care. In recent years, machine learning techniques have shown promise in aiding the prediction of Alzheimer's disease by leveraging the analysis of diverse data sources, including demographic information, genetic markers, biomarkers, neuroimaging data, and cognitive assessments. In this paper, a combined technique called Recursive feature elimination with Artificial Neural Network (R-ANN) is introduced for fast processing and better prediction of AD. In the combined technique, the Recursive Feature Elimination (RFE) selects the most important features in the dataset for reducing the computational complexity and improves the disease prediction accuracy. RFE can handle non-linear relationships between features and the target variable by utilizing machine learning models capable of capturing complex interactions. This allows RFE to potentially capture non-linear dependencies between features and the disease being predicted. The selected features are used in ANN for prediction of AD. The empirical results prove that the accuracy, precision, recall and F-measure of proposed R-ANN is 4.71%, 4.6%, 3.41%and 3.37% better than Regularized Extreme Learning Machine (RELM)technique respectively for AD prediction.