Abstract—‘Smart Homes’ or ‘Intelligent Homes’ are capable in making smart or rational decisions and increase home automation. This is done to maximize inhabitant comfort and minimize operation cost. Tracing and predicting the mobility patterns and usages of devices by the inhabitant, sets a step towards the objective. The paper discusses in detail, the role of certain Prediction algorithms to bring about next event recognition. Further, an Episode Discovery helps in finding the frequency of occurrence of these events and targeting the particular events for automation. The effectiveness of the Prediction algorithms used is demonstrated ;making it clear how they prove to be a key component in the efficient implementation of a Smart Home architecture.
Index Terms—Active LeZi, event history, LZ78, Smart Home, trie.
Aditi Dixit and Anjali Naikare are with Cummins College of Engineering For Women, Pune, India (e-mail: aditi.dxt@gmail.com).
Cite: Aditi Dixit and Anjali Naik, "Use of Prediction Algorithms in Smart Homes," International Journal of Machine Learning and Computing vol.4, no. 2, pp. 157-162, 2014.