Feature Selection and Classification for Searching Light at Night Exposure and Students’ Weight Relationship

Circadian rhythm is essential for living beings. This rhythm regulates sleeping and waking patterns, hormone production, eating habits, digestion, and body temperature in humans and other animals. Prolonged use of Light at Night (LatN) has become a factor that can confuse biological clocks. In our research, we utilized ANOVA to perform feature selection and classification algorithms to explore the potential correlation between night-time light exposure and medical surgeon students’ weight. Our findings revealed that incorporating LatN exposure data resulted in enhanced classification outcomes. With this additional information, and not including BMI and height as features, the classifier was better equipped to distinguish between those who are overweight/obese compared to when only the students’ weight was considered.