# Increasing accuracy of classification physical activity based on smartphone using ensemble logistic regression with boosting method
> Lawi A.

URL kanonis: https://discover.unhas.ac.id/publications/increasing-accuracy-of-classification-physical-activity-based-on-smartphone-usin
Jurnal / Konferensi: Journal of Physics Conference Series
Tahun terbit: 2019
DOI: https://doi.org/10.1088/1742-6596/1341/4/042002
ISSN: 17426588
Citations: 4

## Authors
- Lawi A.

## Abstract
Abstract At present, the smartphone is equipped with several sensors such as Accelerometer, Gravity sensor, and Gyroscope which can be used to recognize human physical activities such as walking upstair and walking downstairs etc. Machine learning is needed to group data and get information. Statistical methods have poor performance in classifying because procedures must be met. To overcome this, an ensemble technique was used. This study proposes the application of the gradientboost ensemble method to classify walking upstair and walking downstairs. The Android-based.apk system is designed for data retrieval using a smartphone. Then, the dataset will be partitioned into 70% training data and 30% test data. The results show that the performance of the ensemble boosting method produces 81.82% accuracy, 86.11% sensitivity and 77.50% specificity.

## Keywords
- Boosting (machine learning)
- Computer science
- Accelerometer
- Gyroscope
- Artificial intelligence
- Ensemble learning
- Gradient boosting
- Logistic regression
- Machine learning
- Pattern recognition (psychology)
- Random forest
- Engineering
- Operating system
- Aerospace engineering

---
Sumber: Discover Unhas — RIMS Universitas Hasanuddin.
Saat mengutip, gunakan DOI bila tersedia atau URL kanonis di atas.