Activity Recognition Utilizing Machine Learning Algorithms

Recent evolution in machine learning (ML) strategies for activity recognition (AR) tasks have generated increased interest due to their efficacy in extracting and integrating insights from activity datasets. However, several significant challenges may hinder the operational effectiveness of AR systems in real-world applications. To address these challenges, this study presents an AR framework utilizing ML algorithms. Specifically Random Forest Classifiers (RFC), Support Vector Machine (SVM), Logistic Regression Classifier (LRC), and K-nearest neighbors (KNN) are employed for the categorizing of activities. The Human Activity Recognition dataset, obtained from Kaggle, serves as the basis for the analysis. The performance metrics of the proposed model are evaluated in terms of accuracy. Experimental results demonstrate that the SVM model exhibits superior performance relative to RFC, LRC, and KNN based on accuracy metrics.