Today machine learning is used widely across all fields of science and technology. One of the most important among them is Health care sector. Health care sector is seeing a drastic and dynamic change post covid-19. Today, the need to visit and consult a doctor is reducing as many hospitals are providing online consultation. Thus, preliminary diagnosis and medication are provided to the patient without physically visiting the doctor. Classification and prediction are supervised machine learning techniques which have wide applications in health care sector such as preliminary diagnosis, quick medication and cost savings. This paper gives insights and results on three machine learning algorithms which classify and predict the disease/prognosis a patient may have based upon the given symptoms. Labelled Dataset is used for classification and prediction of disease based upon symptoms. This dataset is labelled based upon the binary values for each symptom. This system is used by end users like patients who will provide their symptoms and then will receive the most accurate prognosis. This paper also aims to provide a comparative study on 3 different ml classification algorithms based upon their results. The final prediction of the diagnosis is obtained from the cumulative results of 3 algorithms. Hence a more accurate and correct diagnosis is provided to customers in very cost -effective way.