A feasibility study for ambient intelligence based on Arduino and server structure

This paper describes machine learning application in ambient intelligence. The traditional regression analysis in ambient intelligence is based on the mathematical model. While the inductive learning algorithm proposed in this paper is based on big data. The inductive learning algorithm used in ambient intelligence takes the Candidate-Elimination algorithm and Rote-Learner algorithm core ideas as references. Therefore, the output of inductive learning algorithm will be more flexible and accurate than regression analysis. This inductive learning algorithm has been tested in a simple ambient intelligence system based on Arduino and server structure and it performs well.Master of Science (Communications Engineering