Cancer is considered the worst of all diseases. It is a category of diseases that enable irregular growth that may enter or spread to certain body areas. These contrast with healthy, not multiplying tumors. There are 100 different cancer forms that impact humans. With the emergence of machine learning (ML), its uses have been identified in many fields, particularly medical research. It also used for cancer detection when a correct dataset is available. This paper suggests a category boosting (CatBoost) ML algorithm for predicting the different stages of breast cancer, facilitating early diagnosis. The proposed CatBoost algorithm is an efficient method to train and test the available data. To show the CatBoost method's efficacy a detailed comparative analysis has been carried out with other prominent ML approaches. It has been established that the CatBoost is accurate compared to the other ML methods.

Keywords: Breast cancer, Category boosting, Machine learning, Prediction model