In the era of rapidly changing lifestyles with advancement, many people are facing challenges due to mental health disorders such as anxiety, stress, and depression. Due to work pressure, imbalanced life and no social support, the different age groups face different emotional challenges. To support people suffering from mental health disorders, it is required to monitor and track emotions with IoT wearable devices, psychological analysis, and self-assessment surveys. To resolve the mental health issues in working profession with accurate decisions, the traditional machine learning algorithm is not sufficient. There is a need for an ensemble machine learning technique to provide better results. Several studies have been conducted in predicting mental health and supporting emotional challenges through machine learning techniques. In the field of Artificial Intelligence, ensemble techniques are very promising in the mental healthcare sector, as the prediction and diagnosis of mental health disorders can be improved. To optimise the process of decision making, various impactful features of the survey dataset, such as depression, sleep duration, suicidal thoughts, financial stress, and family history, have been closely monitored, which helps in training the model and generating precise decisions. This paper presents the evaluation of machine learning approaches for mental health prediction among working professionals. The variety of machine learning models is gauged through performance metrics with mental health survey data. The research also identifies an efficient model to diagnose mental disorders at an early stage, which may help patients to take necessary action and consult practitioners.