Scholarlink

Diet plays a crucial role in maintaining overall health and managing chronic diseases such as PCOD, diabetes, nephrolith and heart conditions. Despite its significance, conventional dietary recommendations often fail to address personalized needs based on individual health profiles, limiting their effectiveness. To address this, we propose a diet recommendation system powered by transformer networks, achieving an accuracy of 93%. This system provides personalized diet plans in the form of weekly recipes for breakfast, lunch, snacks, and dinner. Recommendations are customized based on user inputs, including height, weight, gender, and health conditions. For individuals without specific diseases, diet suggestions are generated based on BMI calculations. The platform allows users to create accounts using login credentials (username, email, phone number, password) and track their weekly diet adherence. A scoring mechanism evaluates user progress, enabling future dietary adjustments. Additionally, non-logged-in users can access diet predictions. By integrating transformer networks, our system bridges the gap between generalized dietary guidelines and personalized nutrition, offering personalized recommendations for both men and women. Future work emerging to incorporate AI driven recommendation diet based on ingredients.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Transformer Network Based Deep Learning Approach in Customised Dynamic Dietary Endorsements for Chronic Diseases

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • G. Vijayalakshmi,
  • V. Prasanna Kumari

摘要

Diet plays a crucial role in maintaining overall health and managing chronic diseases such as PCOD, diabetes, nephrolith and heart conditions. Despite its significance, conventional dietary recommendations often fail to address personalized needs based on individual health profiles, limiting their effectiveness. To address this, we propose a diet recommendation system powered by transformer networks, achieving an accuracy of 93%. This system provides personalized diet plans in the form of weekly recipes for breakfast, lunch, snacks, and dinner. Recommendations are customized based on user inputs, including height, weight, gender, and health conditions. For individuals without specific diseases, diet suggestions are generated based on BMI calculations. The platform allows users to create accounts using login credentials (username, email, phone number, password) and track their weekly diet adherence. A scoring mechanism evaluates user progress, enabling future dietary adjustments. Additionally, non-logged-in users can access diet predictions. By integrating transformer networks, our system bridges the gap between generalized dietary guidelines and personalized nutrition, offering personalized recommendations for both men and women. Future work emerging to incorporate AI driven recommendation diet based on ingredients.