A Sustainable Artificial Intelligence Framework for Early Identification of Eating Disorder Risk Factors

¹ St Thomas College of Engineering & Technology, Kolkata, West Bengal, India
² Techno India University, West Bengal, India
³ Shri Ramkrishna Institute of Medical Sciences and Sanaka Hospitals
⁴ Technische Hochschule Ingolstadt, Germany

^* Corresponding author : This email address is being protected from spambots. You need JavaScript enabled to view it.

Abstract

The eating disorders are a serious psychological issue that can scarcely be identified unless it causes serious psychological and physical complications. In fact, it is quite significant to detect risk factors as early as possible to have the opportunity to react and control the situation before it worsens. The proposed paper recommends a sustainable AI-based model of enhancing the early detection of the risk factors of eating disorders, which may apply to the mental health and clinical decision-making sphere. The model puts to use different behaviour, cognitions, and background issues with the aim of determining the patterns in relation to emotional eating, body concerns issues, and social pressures. It uses both machine learning systems that consider a significant number of possibilities to make higher quality insights. By using pre-processing methods like categorical encoding, refining features and class balancing, fairness was achieved and predictive reliability was also increased. The model was tested in terms of its robustness with the technique of stratified cross-validation. Moreover, SHAP values were used to visualize the factors that have the greatest impact on the risk of disordered eating. Implementation in clinical practice, learning institutions, and community-based mental health programs are depicted in the framework as having great capabilities of implementation by the inclusion of predictive accuracy, transparency, and sustainability. The outcomes emphasize the significance of understandable and sustainable AI as a proactive tool of early screenings and risk assessment, which contributes to the improvement of the outcome of biomedical and healthcare analytics.