Published Paper Details:

The Internet of Things (IOT) is an emerging research area promising many interesting solutions to various problems encountered in various domains. Smart homes applications is one such branch that evolve from IOT with the huge challenges of data storage and handling. Multiple activity recognition is the major challenge in smart homes application that incorporates multiclass learning. The efficiency of ensemble learner in handling multiclass problem and collective dissicion delivered prompt its uses in the smart homes application In this paper we deal with multiple activity recognition problem on various ensemble learner including bagging ,boosting and random forest. The standard van Kasteren dataset contains three housedata with eight activities of thirty days. We perform our experiment on the pre-processed collected data and applied six learners i.e. three from individual learner and three from ensemble learners. On performing the extensive experiment it was found that the group of ensemble learner outcast the simple learners