Stacking/Ensemble Model for Smartphone-Based Human Activity Recognition using Feature Engineering

Published: 19 Mar 2024, Last Modified: 29 Apr 2024Tiny Papers @ ICLR 2024 ArchiveEveryoneRevisionsBibTeXCC BY 4.0
Keywords: HAR, Tracking, Ensemble, Health, Human Activity Recognition, Maximum Relevance Minium Redundancy
TL;DR: Our project aims to create a model that accurately recognizes human activity using data from smartphones, utilizing ensemble models and the Maximum Relevance - Minimum Redundancy feature engineering technique.
Abstract: Human activity recognition, commonly known as HAR, involves identifying numerous physical activities individuals conduct across diverse situations. Examples of such actions include walking, jogging, running, ascending stairs, descending stairs, and more. Our work attempts to construct a model that accurately recognizes human activity based on data obtained from cellphones. To achieve optimal performance, we plan to use an ensemble model in conjunction with feature engineering.
Submission Number: 80
Loading