Elderly Home Meal Management System

Abstract

Through talking to a frontline caregiver who had more than 15 years of experience working in both well-established and private elderly home, it was understood that some private elderly homes had not been able to offer comprehensive services; for instance, without providing nutritional meals for the elders. Therefore, this had inspired me to train a model that aimed to provide elders at a private elderly home with the right meal plans based on their health conditions. The model was trained using a Decision Tree Classifier (DTC), Python, and a dataset with key parameters of elders' systolic and diastolic blood pressure, heart rate, blood sugar level, allergic food and whether the elder had chewing difficulties. The dataset consisted of 448 rows of data, that was created based on normal reference ranges of the mentioned health parameters and was distributed normally. The recipe was the element to be predicted by the model, by also referencing to the recipes' tags as stored in the .csv file.