A Machine Learning Implementation for Mental Health Care. Application: Smart Watch for Depression Detection

Machine learning is one of the most significant dimensions of artificial Intelligence. It is being used in almost all fields of science and technology. The Healthcare sector is one such realm where the application of machine learning has given excellent results. Moreover, in combination with the Internet of Things (IoT), machine learning has been widely successful in the healthcare sector. Still, there are some areas that remain devoid of the growing technology. Mental illness is one of the areas where there hasn’t been any perfect treatment. Predicting whether a person has a mental illness itself is the big challenge. Psychologists provide assessment and therapy to their clients with one-on-one physical interactions. Still, there is some ambiguity regarding the treatment. Although psychologists prescribe various medications for their clients like anti-depressants, sleeping pills, etc.; still, the medication hasn’t been able to cure or eradicate the sickness. There may be multiple reasons why a person is going through a certain situation like society, work pressure, family, etc. Our research on this topic will be limited to predicting such sickness in the human body and identifying what the person is going through using the previously recorded dataset. We will be using Logistic Regression, Support Vector Machine (SVM), Decision Tree, K-Nearest Neighbor, and Naïve-Bayes algorithms for creating ensemble models and further compare the models. We have applied the proposed algorithms on the Kaggle dataset having 334 sample sizes with 31 different fields about unemployment and mental illness. In the end, the test result of this application can be an authentic example of IoT in healthcare.