Machine Learning For Disease Biomarker Detection From Genome Expression Data

In light of the absence of an effective cure or treatment for Alzheimer's disease (AD), understanding its underlying genetic mechanisms is of utmost importance. This study introduces a novel machine learning (ML) method to identify potential biomarker genes associated with, utilizing gene expression data from the GSE5281 dataset. Employing computational and statistical techniques, this research aims to contribute to this understanding by proposing a ML framework for biomarker discovery. The study utilises a set of 14 genes that show promise in relation to AD, shedding light on their potential biological significance. By providing key insights into this ML-based approach, the study enriches ongoing endeavors to unravel the intricate genetic landscape of Alzheimer's disease. The dataset, comprising 161 samples (87 AD, 74 controls) with gene expression data and corresponding clinical information, demonstrates classification accuracies ranging from 81.59% to 89.78% for testing and 78.52% to 93.74% for training across various algorithms.