CSC310-fall25/regression_obesityWeight_model

Obesity Weight Prediction Model — Linear Regression

Overview

This model predicts a person’s weight (kg) based on height (m) and age (years) using a Linear Regression model from scikit-learn.

---

Training

Detail	Value
Algorithm	LinearRegression()
Features	Height, Age
Target	Weight
Train/Test Split	75% / 25%
Random State	42
Dataset	ObesityDataSet_raw_and_data_sinhtetic.csv

---

Performance

Metric	Score
MSE (Mean Squared Error)	511.55
R^2 Score	0.2777

These results indicate that height and age alone do not fully explain weight — important factors like diet, genetics, and exercise are missing.

---

Visualization

Below is a scatter plot showing predicted vs true weights:

The wide spread around the regression line shows prediction uncertainty for heavier individuals.

---

Limitations

• Only two features used → reduced explanatory power
• Synthetic dataset — not reflective of real population variation
• Performance not suitable for real-world medical decisions

This model is intended for educational use only.

---

Strengths

• Easy to interpret
• Fast and simple
• Good educational model

Weaknesses

• Low accuracy
• Missing key health variables
• Not production-ready

Citation

• "Estimation of Obesity Levels Based On Eating Habits and Physical Condition ." UCI Machine Learning Repository, 2019, https://doi.org/10.24432/C5H31Z.