Using ML Algorithms for Predicting Breast Cancer

Abstract

The most common disease that can be seen in women is breast cancer. It is about 30% of all new female cancers each year. In 2020, 2.3 million new cases were found globally and about 685,000 deaths occurred due to it. The American Cancer Society estimates about 297,790 new cases in 2023. In most cases, it is found in women with 50 years or more age, but men can get breast cancer too. About 1/100 breast cancers diagnosed in the United States are found in a man having a family history of breast cancer increases risk in both males and females. According to 2021 statistics, it was found that 281,550 new cancer cases were discovered in the United States. Due to the rapid increase in death cases because of breast cancer, there is a need to find an effective solution to this problem. Machine Learning Algorithms can help in providing a solution to this problem by predicting breast cancer at an early stage so that it can be cured at an early stage only, which leads to preventing deaths. In this paper, we have used the following Machine Learning Algorithms like DT (Decision tree), RF (Random Forest) classifier, NB (Naïve Bayes) classifier, KNN (K-Nearest Neighbors) classifier, AdaBoost (Adaptive Boosting), GBDT (Gradient Boosted Decision Trees), SVM (Support Vector Machine), SGD (Stochastic Gradient Descent), RF (Random Forest) classifier. And we have applied feature selection to extract the best attributes so that the ML classifier can provide better accuracy to our model and helps in saving the lives of many people. The accuracy of GDBT is 97%, the SVM classifier is 96.4%, AdaBoost is 96%, SGD is 94%, the RF classifier is 92%, KNN is 90% DT classifier is 90% and the NB classifier is 90%. Out of all GBDT provides the best accuracy, which is 97%.