Human Activity Recognition using Active Learning Methodology

Abstract

In current technology, presenting detailed and exact information about one’s daily activities is the major task in artificial intelligence. This paper represents the multiple classification techniques used to monitor the behaviours of aging people. It can also play an important role in health care monitoring system and surveillance systems. Human Activity Recognition (HAR) dataset is used for evaluating and comparing the prediction accuracy of the dictionary learning algorithm, Naive Bayes and J48 algorithms. Based on the classification, J48 algorithm is superior compared to other classifier algorithms. J48 and Naïve Bayes machine learning algorithms are evaluated on WEKA tool and their efficiency is compared with Dictionary learning algorithm for achieving better results on the given dataset.