Cancer is being considered as the deadliest disease among all diseases across the globe. Since two decades, many academic researchers are contributing their work for automated detection of tumor from the data of screening modalities and the results of those researches are enhancing the probability of earlier detection of the disease which enable for proper timely treatment. This review article does a survey of several machine learning algorithms being involved in the earlier detection and classification of tumors and this review is expected to provide a detailed basis for the researchers who carry out their study in the development of automated diagnosis system for cancer detection.