Development of a Health Prediction and Assistance System using ANN and T5-Small Models
Ashish Nain, Ayush Roy, Abhay Mahajan, Vusthela Kavyasri, Aniket Kailas Shahade, Mayur Gaikwad · 2025
In this work, ANN and the T5-Small model are proposed for constructing a health prediction model and a chatbot for health support. The parameters adopted by the ANN model to determine the health of an individual were based on the daily activities, occupation, sleep pattern, and other activities. Through the ability of ANN to capture complicated patterns and combinations, there is always an improvement of classification. Furthermore, fine-tuning also helped with gaining better insight of the data and the connections between the data, which are considered to be more complex. The system was on a local computer and the Arizona version was on the same machine for testing and modifications. The T5-Small model was used to teach the chat assistant what kind of health problems can occur and when it is better to consult with special consultants. This work provides a useful and effective approach to TPH, focusing on individual, accurate, and useful prediction.