Klasifikasi Gangguan Tidur Berdasarkan Gaya Hidup Menggunakan Algoritma Decision Tree

Abstrak

This study aims to build a classification model for sleep disorders including Sleep Apnea, Insomnia, and Normal based on lifestyle using the Decision Tree algorithm. The secondary dataset contains lifestyle variables and health indicators (such as sleep duration, sleep quality, daily steps, stress level, heart rate, and BMI) with a total of more than 400 entries. The data was preprocessed by converting duration and blood pressure to numeric variables and encoding categorical variables, and divided into training data (80%) and test data (20%) in RapidMiner. The results showed that the model achieved an accuracy of 85.33%, and the precision, recall, and f1-score metrics were around 84%. The most influential variables included daily steps, heart rate, sleep duration, sleep quality, and stress level. The decision tree structure also successfully extracted easily interpretable classification rules, for example, individuals with 5,300 steps and a heart rate of 83.5 bpm tend to be classified as Sleep Apnea. This model has the potential to be used in lifestyle-based sleep health decision support systems. Further research is recommended using larger datasets, additional algorithms, and variables such as caffeine consumption and device use to improveaccuracy.

Identifier

• OpenAlex: https://openalex.org/W7166719887
• DOI: https://doi.org/10.26905/jisad.v3i2.16038