Depression is a psychological well-being disorder. As per WHO, it has been assessed that generally 5.0% of grown-ups experience the ill effects of depression. Depression is likewise known to be one of the main sources of handicap around the world. The death toll has been increasing year after year. It has been found that there are various treatments for mild, moderate and severe depression, but it’s hard to detect and determine the severity of depression accurately. Hence, this system will use diagnostic data from PHQ-9(Patient health questionnaire) and determine accuracy of the output. Since PHQ-9 is the standardized questionnaire used by doctors and hospitals as the first step to determine the severity of depression in a patient, this system will be using four machine learning models: KNN, Logistic regression, Decision Tree, and SVM and then cross-validating those results to determine the accuracy of PHQ-9 results. These learning models are chosen as currently they are the best type of regression models for depression detection, a similarity which can be observed in heart disease classification. The conducted experiments show statistical varied data which is statistically more accurate than previous works.