Hybrid Deep Learning and Ensemble Methods for Alzheimer’s Detection from EEG Signals

Alzheimer’s disease (AD) is a progressive neurological disorder characterised by aberrant behaviour, memory loss, and cognitive impairment. Electroencephalography (EEG) is an efficient method that provides valuable information on brain activity and can be used to predict AD. The application of EEG data for AD classification is examined in this abstract, with particular attention paid to the study of frequency bands, and Machine Learning (ML) techniques. This algorithms classifies AD with excellent levels of accuracy, indicating possibility of EEG-based methods for early disease identification. In order to classify AD in its early stages, this research proposes a unique deep learning model based on convolutional neural networks (CNNs) and signal processing techniques. The proposed model aims to classify patients into two categories: AD positive and negative. At the early stage the dataset was preprocessed using techniques such as label encoding and noise removal. Empirical mode decomposition (EMD) and variational mode decomposition (VMD) are the techniques applied in signal processing. The classification is done using Deep Learning (DL) and Machine Learning (ML) methods are used to develop the best model. ML algorithms namely Stacking classifier, Gradient Boosting, Stacknet, AdaBoost, and DL algorithms like Convolutional Neural Networks (CNN) and Deep Neural Network (DNN) architectures have shown promise in AD classification. Random Forest and AdaBoost excel in ensemble learning, while CNN and DNN models capture intricate patterns in EEG data, further enhancing the accuracy of diagnosis. In conclusion, EEG data offer non-invasive and economical method for classification of AD and contain useful information.