抽象的

Intelligent Machine Learning System For Smart Room Using Sensor Network

C.Sugitha, Fleena Christy, S.Chandrasekaran

Advent development in the Micro and Nano electronics paves way for sophisticated living of human using sensors. In Smart environment, human interaction with the computing system is not manual commands, the need of human is predicted by the system by means of data collected through sensors, thus human can enjoy automated sophisticated living environment. The need of human is detected, analyzed and act based on the information collected by sensors. Particularly the need is predicted only by recognizing the human activity. The human activity is recognized by means of efficient machine learning system using data collected by ambient sensors, fixed in all daily usable things and wearable sensors worn by the human who is under monitoring. The aim of creating this prototype is to develop automated machine learning system that does not need any manual interpretation for decision making. Thus the proposed machine learning system is based on Adjustable Fuzzy Clustering (AFC) algorithm, for data grouping to ease the job of classification system and Fuzzy Neural Network (FNN), for supporting incremental learning. This proposed system resembles human intelligence in making accurate decision even in uncertain situation. By the implementation of this system shows higher accuracy of recognizing human activity even with large data from different sensors.