According to the statistics, the disease with which most of the women die is breast cancer. Lot of new cases and deaths because of this cancer places this disease as a major public health issue. The diagnosis of this disease at the starting stages will help the treatment by which the mortality can be reduced. This leads to extensive research in the diagnosis and classification of patients based on its malignancy. Lot of machine learning algorithms was used to diagnose this disease. This work analyses the various works done in this area. Also it shows the comparative study of those algorithms.