Automated Machine Learning based Elderly Fall Detection Classification

Firdous Kausar*, Medhat Awadalla, Mostefa Mesbah, Taif AlBadi

*المؤلف المقابل لهذا العمل

نتاج البحث: المساهمة في مجلةArticleمراجعة النظراء

1 اقتباس (Scopus)

ملخص

As we grow older, one of the most concerning aspects of our lives becomes increasingly challenging to manage our health. Fall is a leading cause of health problems or death in the elderly population. Using a wearable sensors device, this research presents a strategy for identifying and distinguishing fall activities from activities of daily living (ADL) in older persons. The conventional Machin learning method was applied by extracting features from telemetry data after pre-processing, and feature extraction. It is then compared to non-coding Automated Machine Learning (AutoML) method, where all the selected classifiers get automatically optimized. Furthermore, machine learning algorithms such as Support Vector Machine, K-Nearest Neighbor, Random Forest tree, and Artificial Neural Network are used to categorize acceleration signals as falling or regular activity. The test results indicate that AutoML can predict exceptionally accurate results in binary classifications with 99.9% accuracy on three of the four machine learning techniques it was tested against.

اللغة الأصليةEnglish
الصفحات (من إلى)16-23
عدد الصفحات8
دوريةProcedia Computer Science
مستوى الصوت203
المعرِّفات الرقمية للأشياء
حالة النشرPublished - يناير 1 2022
الحدث17th International Conference on Future Networks and Communications / 19th International Conference on Mobile Systems and Pervasive Computing / 12th International Conference on Sustainable Energy Information Technology, FNC/MobiSPC/SEIT 2022 - Niagara Falls, Canada
المدة: أغسطس ٩ ٢٠٢٢أغسطس ١١ ٢٠٢٢

ASJC Scopus subject areas

  • ???subjectarea.asjc.1700.1700???

قم بذكر هذا