Human Activity Classification Approach on Smartphone Using Monkey Search Algorithm

Human activity recognition represents a big step in health care and monitoring. The assessment of studying physical activity became has a great value in health care and monitor human activity patterns. Classification of human activities started to be possible with modern technologies such as smartphone. The current methodology proposed a model to classify human activity patterns based on smartphone data and optimization technique. Previous studies used traditional approaches to recognize human activity that rely on machine learning techniques. The previous studies achieved acceptable accuracy with simple activity and poor accuracy with complex activities. In this paper, a proposed approach depended on traditional classifiers and monkey algorithm as an optimization technique to classify complex or simple activities The monkey algorithm is adapted to be used as a classification algorithm. The experimental results shows that the classification model achieved a high performance in classifying simple activities as previous studies with accuracy up to 93.4%. Also, the proposed model provides an acceptable accuracy above 71.8 % that is better than previous studies. Finally, we put out some future research points in studying human activity and optimization algorithms.