Page 68 - AIH-2-4

Artificial Intelligence in Health Synthetic data for obesity level prediction

Figure 21. Obesity class distributions by gender on height and weight axes

Figure 22. Associations between gender, calorie tracking, and obesity level on the age and weight axes

3.3. Data preprocessing and ML algorithms into ML algorithms. Among the attributes subjected
to one-hot encoding, the gender variable was coded
Given that the majority of the attributes in the dataset are as male = 0 and female = 1. For binary attributes with
categorical, both one-hot encoding and label encoding “yes”/“no” responses, such as family history of obesity,
methods were applied to prepare the data for input frequent consumption of high-calorie foods, smoking,

Volume 2 Issue 4 (2025) 62 doi: 10.36922/AIH025140027