Abstract :

Human Activity Recognition (HAR) Has Emerged As A Transformative Technology In Pervasive Computing, Enabling Intelligent Monitoring, Automated Behaviour Analysis, And Personalized Digital Services. Recent Advancements In Wearable Sensors And Machine Learning Have Accelerated The Development Of Robust HAR Systems, Particularly For Applications In Healthcare, Fitness Tracking, Human–computer Interaction, And Assisted Living. However, The Complexity Of Human Motion Patterns, Variability In Sensor Noise, And The Challenge Of Generalizing Across Users Continue To Hinder Consistently Accurate Predictions. This Research Presents A Hybrid Machine-learning Framework For Wearablesensor-based HAR That Integrates Temporal Feature Extraction With An Optimized Ensemble Classification Pipeline To Improve Recognition Accuracy In Real-time Environments. The Proposed Model Processes Tri-axial Accelerometer And Gyroscope Data, Performs Multi-stage Preprocessing, And Applies Synthetic Feature Fusion To Enhance Discriminative Capability. Evaluation On Benchmark Datasets Demonstrates Significant Performance Improvement Compared With Conventional Single-model Approaches. The Study Also Highlights The Impact Of Sampling Frequency, Window Size, And Sensor Placement On Classification Reliability. The Outcomes Indicate That Hybrid Models Effectively Address Limitations Related To Feature Redundancy And Activity Overlap, Thereby Enabling HAR Systems To Perform Reliably Across Heterogeneous User Groups. The Research Contributes A Scalable Methodology Suitable For Deployment In Wearable IoT Devices And Mobile Platforms, Offering High Efficiency And Adaptability For Real-world Applications. Keywords: Human Activity Recognition, Wearable Sensors, Machine Learning, Feature Fusion, Ensemble Learning, Accelerometer, Real-Time Classification

Published:

12-12-2025

Issue:

Vol. 25 No. 12 (2025)

Page Nos:

216-225

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite