“Advanced Anxiety Risk Stratification Using Regularized Deep Neural Networks”

Anxiety disorders are widespread and significantly impact daily life, necessitating accurate and timely assessment for effective intervention. Traditional methods of anxiety evaluation are often subjective and time-consuming. This study introduces a deep neural network (DNN) model with L2 regularization designed to classify anxiety into four categories: low, moderate, high, and no. For stable learning, the model employs batch normalization, dropout for regularization, and hidden layers with rectified linear unit (Re functions. Trained on a diverse Kaggle dataset, which includes psychological and physiological features, the DNN achieves a remarkable classification accuracy of 92%. Anxiety evaluations may be automated using deep learning techniques, as demonstrated in this paper, providing mental health professionals with an effective tool for developing customized treatment programs. By adding more data sources, future research attempts to improve the model's predicted accuracy and scalability, thereby advancing automated anxiety assessment in clinical settings.