Across the world, cancer becomes a catastrophe for a human being who is suffering from it. Cancer can be diagnosed at a premature stage to overcome the consequences at a later stage and the possibility of endurance considerably, as it can support appropriate medical action to patients. One of the frequently used innovative technologies for the diagnosis and detection of cancer is Machine learning (ML). In recent times ML has been used for the prediction and prognosis of cancer. Machine learning enables the creation of algorithms that can learn and make predictions. Various Machine Learning techniques can build a model to diagnose cancer based on finding accuracy level. It is possible for early detection of cancer through machine learning where we train the machine with previous data. This paper aims to predict cancer type based on symptoms given by the user. Here we adopted a supervised learning algorithm and then use the Logistic Regression based on accuracy and recall score i.e., the algorithm which gives high accuracy level and recall score. The proposed System executes with good performance as it generates accurate results.