Classification of Human Blood Cells Based on Subtypes Using Linear Discriminant Analysis on Microscope Images

This study aims to develop an automated classification system for human blood cells using microscopic images to improve the accuracy of subtype identification. To overcome the limitations of manual classification, the research adopts an artificial intelligence approach using the Linear Discriminant Analysis (LDA) algorithm, chosen for its effectiveness in dimensionality reduction and data group separation. The study follows the Knowledge Discovery in Databases (KDD) methodology, involving data selection, preprocessing (normalization, enhancement, segmentation, and noise removal), feature extraction using the Gray Level Co-occurrence Matrix (GLCM), and classification into eight blood cell types. The model's performance is evaluated using metrics such as accuracy, precision, recall, and F1-score. The research aims to contribute to more efficient and accurate medical image classification systems and demonstrate the potential of LDA in AI-based medical applications.