Machine Learning Algorithms for Analysis and Anticipation of Breast Cancer: A Comparative Approach

In the medical domain, data analytics play a critical role in diagnosis and treatment. Huge amount of data should be analyzed by using machine learning techniques to generate different tools for prediction and categorization, allowing practitioners to make better decisions. The clustering and classification technique is used to predict the occurrence of Breast Cancer (BC). In order to achieve accurate predictions, the system uses the Wisconsin Original Breast Cancer Dataset obtained from the UCI Repository. The datasets are analyzed in R Studio, where the records are grouped by using the Partition based clustering and further categorized by using Decision Tree and Support Vector Machine (SVM) algorithms. The proposed model predicts whether a tumor is benign (noncancerous) or malignant (cancerous) and the model is then evaluated by using N-fold cross validation. The main aim of the proposed system is to demonstrate the algorithms that are optimal for performing healthcare-based prediction tasks. The accuracy rate, as well as the efficiency and efficacy of each method are then determined. The experimental findings show that the proposed approach has the highest accuracy with 90.34 percent accuracy for the Decision Tree Classifier and 93.98 percent accuracy for the Support Vector Machine. Further, the Support Vector Machine (SVM) is considered as one of the most suitable classifiers with a 3.64 percent improvement in accuracy.