Comparison of Dense Net and over Logistic Regression in Predicting Leukemia Classification with Improved Accuracy

Abstract

This study compares the performance of densenet and support vector machines (SVMs) in the diagnosis of leukemia disease, with the aim of improving the accuracy of the classification results. Materials and Method The Kaggle website is where the dataset was found. The dataset consists of 20 samples per group in JPG files with a resolution of 96 dpi and 512×512 pixel size.The sample size is determined using a pretest power of 80%, a threshold of 0.05, and a confidence interval of 95%. Results: For leukemia, dense net is 96.5%, whereas logistic regression is 89%. The significance levels for Densenet and logistic regression are data with p=.000 (p<0.05) statistical significance difference respectively. Conclusion: Based on the findings, I believe that densenet performs superior to logistic regression.

Keywords

• novel-accuracy
• dense net
• disease
• leukemia
• logistic regression
• machine learning
• red blood cells
• white blood cells