Classifying Body Type based on Eating Habits and Physical Condition using Decision Tree Technique

• Published in: IVIT 2022 - Proceedings of 1st International Visualization, Informatics and Technology Conference
• Main Author: Bahrin U.F.; Jantan H.; Sani Mohd Sofian M.A.; Syahirah Ismail I.; Aishah Samsudin S.H.
• Format: Conference paper
• Language: English
• Published: Institute of Electrical and Electronics Engineers Inc. 2022
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85148633200&doi=10.1109%2fIVIT55443.2022.10033335&partnerID=40&md5=502bc6ea9f25de230c0d017592002f43

Nowadays, due to busy schedules, many people are unaware of what they are eating and their physical condition. This scenario will lead to various health issues such as obesity, diabetes, blood pressure, etc. Hence, it has become very essential for people to have a good balanced nutritional healthy diet to deal with those issues. Therefore, it is important to determine what factors may be conducive to healthy eating behaviors among people with different Body Mass Index (BMI). A predictive analysis approach in data mining can be used to identify the food consumption pattern in people's eating habits and how it is related to their body type. This study aims to classify body types based on eating habits and physical conditions using a decision tree induction algorithm. Several phases have been conducted in this study such as data understanding, data preparation, modeling, and evaluation. In the experimental phase, the datasets that are known as full dataset and reduced dataset have been used to identify which dataset will produce high accuracy. As a result, it is shown that a full dataset produces higher accuracy compared to a reduced dataset. Perhaps there is room for improvement in the reduced dataset by applying other attribute selection methods to produce better accuracy of the classifier. This study brings a high significance for effectiveness and efficiency in eating habits and physical condition analysis based on body type, and it can also be explored for other classification methods for future work enhancement. © 2022 IEEE.