Abstract: One of the most in demand research topic in today’s technology world is medical area and cancer is one of them. The second main cause of death in the world is cancer. In 2015 about 8.8 million people have died due to cancer [5]. For the early detection of breast cancer several types of research have been done to start the treatment and increase the survivability. Breast cancer affects the women mentally as well as emotionally. The goal of this challenge is to provide a framework that uses a cancer medical dataset as input and then analyses the dataset to produce findings that help medical experts better understand the state of the disease. The majority of studies concentrate on mammography results. However, incorrect detection in mammography pictures can occasionally result, endangering the patient’s health. Cancer that forms in the cells of breast is said to be breast cancer. The cancer should be cured if it is diagnosed in early stage. So, with the help of machine learning algorithm the cancer should be diagnosed early. Most of the women lives are affected by the breast cancer in all over the world.

There are two types of tumors that can be found in breast cancer i.e. malignant or benign. If a person is having a cancer disease, then it will be categorized as malignant otherwise it is known as benign. In 2020, 685000 deaths and 2.3 million women diagnosed with breast cancer globally, somewhere in world in every 14 seconds, a woman is diagnosed with breast cancer [1]. Patient life from the breast cancer can be saved only if it is found in early stage; if it is diagnosed later then the chances of survival are less. If the cancer is diagnosed early then the patient will get a better treatment. This study will concentrate on a few machine learning methods for identifying if a breast cancer is malignant or benign. The Wisconsin Breast Cancer Dataset, which was acquired via Kaggle, was used in this study. Our goal is to evaluate how accurately various machine learning algorithms can detect breast cancer. These include Random Forest Classifier, Decision Tree Classifier, Support Vector Machine, and K-Nearest Neighbors. All the experiments are conducted on a Jupiter platform. After the analyzing the accuracy of each algorithm the most suitable one is Support Vector Machine that gives the better accuracy among all i.e., 98%.