Easy Ensemble: Simple Deep Ensemble Learning for Sensor-Based Human Activity Recognition

Sensor-based human activity recognition (HAR) is a paramount technology in the Internet of Things services. HAR using representation learning, which automatically learns feature from raw data, mainstream method because it difficult to interpret relevant information sensor data design meaningful features. Ensemble learning robust approach improve generalization performance; however, deep ensemble requires various procedures, such as partitioning and training multiple models, are time-consuming computationally expensive. In this study, we propose an easy (EE) for HAR, enables implementation single model. addition, techniques (input variationer, stepwise ensemble, channel shuffle) EE. Experiments on benchmark set demonstrated effectiveness EE their characteristics compared with conventional methods.