UCLA

This article examines the efficacy of machine learning techniques in analyzing brain tumor MRI images with the aim of reducing the workload of medical professionals. The study compared various machine learning methods for processing MRI data and their accuracy. Results show that convolutional neural networks (CNN), including Custom CNN, ResNet v2, and VGG 16, outperform traditional machine learning algorithms such as random forest, K-NN, and SVM in tumor classification accuracy. VGG 16 shows the highest accuracy, reaching 98.73%, and has the smallest loss compared to other CNN models. These data results provide insights into the comparative performance of machine learning models, revealing their strengths and limitations in processing different brain tumor images.