Prediction of Trait Anxiety in Humans

Concerns about mental health have been increasingly common in recent decades. Mental health problems have existed from the beginning but have always been stigmatized by people. According to statistics, one out of every four persons in the world today has a mental health problem and despite this, only a few people receive mental health assessment or treatment. Anxiety is one of the most common psychological emotions felt and is often stated to be healthy but becomes a medical disorder when a person regularly feels disproportionate levels of it. The aim of this study is to predict the trait anxiety in humans at an early stage so that it can be acted upon accordingly by professionals and lead to a better diagnosis and treatment of the subject. We have implemented machine learning algorithms to analyze data and produce a highly accurate model. We have used the TMAS or Taylor Manifest Anxiety Scale Response as dataset, which is a response questionnaire that consists of 50 True and False questiossns and predicts anxiety in humans based on the test score obtained by the subject. The test is scientifically backed and used extensively in the field of psychiatry bearing positive outcomes. The machine learning algorithms used were support vector machine (SVM), random forest (RFA), and K-nearest neighbor (KNN). Highest accuracy was found by using the support vector machine (SVM) algorithm.