Article Per Year (5 Year)
p-Index From 2020 - 2025
0.23
P-Index
This Author published in this journals
All Journal Journal of Multidisciplinary Inquiry in Science, Technology and Educational Research
Hidayat Lubis, Fitra
Unknown Affiliation
Humanities Economics, Econometrics & Finance Education Social Sciences Other

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
Penerapan Multi-Layer Perceptron untuk Prediksi Durasi Tidur Berdasarkan Faktor Kebiasaan Harian Risdi Aulia, Rafif; Hidayat Lubis, Fitra; Sofinah Harahap, Lailan
Journal of Multidisciplinary Inquiry in Science, Technology and Educational Research Vol. 2 No. 1 (2025): NOVEMBER 2024 - JANUARI 2025
Publisher : UNIVERSITAS SERAMBI MEKKAH

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32672/mister.v2i1.2326

Abstract

This study applies a Multi Layer Perceptron (MLP), a type of Artificial Neural Network (ANN), to predict sleep duration based on daily habits, including screen time, exercise, and caffeine intake. The methodology involves data preprocessing, MLP architecture design, hyperparameter tuning using Grid Search, and model evaluation. The final model configuration includes two hidden layers with 10 neurons each, utilizing the tanh activation function and adam optimizer with a learning rate of 0.1. The model evaluation on test data shows promising accuracy, with a Mean Squared Error (MSE) of 0.065 and Mean Absolute Error (MAE) of 0.204. These results indicate that the MLP model effectively captures complex patterns in the dataset and provides accurate sleep duration predictions. However, certain samples showed significant prediction discrepancies, suggesting the potential influence of unobserved factors, such as health conditions or stress. Further research could improve model performance by including additional features or exploring alternative models like Random Forest or Gradient Boosting.
Co-Authors Risdi Aulia, Rafif Sofinah Harahap, Lailan