WBCs, or white blood cells, Leukemia Excessive leukocyte production in the bone marrow is the etiology of leukemia, and image-based detection of malignant WBCs is crucial for its identification. For the identification of malignant WBCs, it is essential to accurately classify them from medical imaging. The accuracy and computing efficiency of conventional techniques, including decision trees and naive Bayes classifiers, are constrained. This study offers a sophisticated approach to WBC leukemia classification by using an Improved Salp Swarm Algorithm (ISSA) for feature selection and the Grey Level Co-occurrence Matrix (GLCM) for feature extraction. The most pertinent features are chosen by this bio-inspired optimization technique, which also eliminates noisy and highly correlated features. A deep neural network model that has been modified (EDLM) is intended to improve prediction accuracy and avoid overfitting. used theThe suggested strategy outperformed traditional techniques like Naivenaive Bayes (74%) and SVM (89%), achieving a classification accuracy of 95% when applied to a public WBC Leukemia reference dataset. The suggested method shows promise as a useful tool for WBC leukemia classification, providing both high accuracy and computational efficiency, with high precision and recall values of 94% and 88%, respectively.