A Comparative Machine Learning Approach for Smartphone – Based Human Activity Recognition Using Feature Engineering | IJCT Volume 13 – Issue 2 | IJCT-V13I2P113
International Journal of Computer Techniques
ISSN 2394-2231
Author
Aditya Pratap Yadav, Meidingu Lourembam, Suhas Achappa, Achyut N Gowda, Prasanth T
Abstract
This Abstract Human Activity Recognition (HAR) has become an essential study area in applications such as medical care observation, fitness monitoring, and smart environments. With the rising availability of smartphone sensors, scalable and real-time activity classification has become practical and workable. This paper presents a comparative analysis of four supervised machine learning models; k-Nearest Neighbors (k-NN), Logistic Regression, Random Forest; and Support Vector Machine (SVM), using the UCI HAR dataset. The models are assessed applying cross-validation and numerous and manifold performance metrics. Experimental outcomes show that Random Forest achieves the supreme and paramount accuracy of 93.07% with robust generalization performance. Feature significance analysis discloses that gravity-based and angle-related features considerably affect classification. The study highlights the effectiveness of ensemble approaches for feature-based HAR systems and offers insights for real-world deployment.
Keywords
HAR, Machine Learning, Random Forest, Smartphone Sensors, Classification
Conclusion
This paper presents a comparative analysis of four machine learning models for smartphone-based Human Activity Recognition. The results demonstrate that Random Forest provides the best performance in terms of accuracy and stability. Its ability to handle high-dimensional data and reduce overfitting makes it a suitable choice for real-world applications.
Future work will focus on exploring deep learning approaches, real-time mobile deployment, and personalized activity recognition systems
References
[1]D. Anguita et al., “A Public Domain Dataset for Human Activity Recognition Using Smartphones,” ESANN, 2013. [2]L. Breiman, “Random Forests,” Machine Learning, vol. 45, no. 1, pp.
5–32, 2001. [3]T. Cover and P. Hart, “Nearest Neighbor Pattern Classification,” IEEE Transactions on Information Theory, 1967. [4]C. Cortes and V. Vapnik, “Support Vector Networks,” Machine Learning, 1995. [5]F. Pedregosa et al., “Scikit-learn: Machine Learning in Python,” Journal of Machine Learning Research, 2011. [6]J. R. Kwapisz et al., “Activity Recognition using Cell Phone Accelerometers,” ACM SIGKDD, 2011.
How to Cite This Paper
Aditya Pratap Yadav, Meidingu Lourembam, Suhas Achappa, Achyut N Gowda, Prasanth T (2026). A Comparative Machine Learning Approach for Smartphone-Based Human Activity Recognition Using Feature Engineering. International Journal of Computer Techniques, 13(2). ISSN: 2394-2231.
