Technological advancement in the healthcare industry has been helping people to suggest hospitals and doctors to travel for treatment, where to admit, and which hospitals are the simplest for treating the desired disease. This paper intends to use machine learning (ML) algorithms for smart diagnosis to classify a disease using the patient’s symptoms. People can use this model to predict their illness before seeking medical attention from a specialist doctor. This research focuses on existing techniques applied to human disease diagnosis in the medical field. Three different models are applied to a dataset for this experiment. Experimental results and analysis show that it is possible to construct a self-diagnosis system in the healthcare system to prevent wrong-directional treatment.