Enhancing Malignant Tumor Detection: A PSO-Optimized Multilayer Perceptron Neural Network Method

The development of precise diagnostic techniques is necessary for any cancer detection. in this paper, using the Breast Cancer Dataset and a Multilayer Perceptron Neural Network (MLP-NN), an efficient method for detecting malignant tumors is presented utilizing Particle Swarm Optimization, for setting the number of hidden layers. The proposed model demonstrates promising results in accurately differentiating between benign (۹۸%) and malignant (۹۷%) tumors. For the sake of evaluation, support vector machine, and Logistic Regression, were also used to classify the breast cancer dataset. The results show that the MLP method has the best performance among other classifiers.

کلیدواژه ها:

MLP ، Breast Cancer ، Classification ، SVM ، Logistic Regression