A self-adaptive distributed decision support model for Internet of Things applications

The term of Internet of Things (IoT) is an emerging concept that has already made an impact on many research domains by providing new solutions and ideas. However, we noticed that many IoT applications are more focused on the communication' part and are still relatively weak in intelligent' aspects. Consequently, in this paper we proposed an approach for a self-adaptive distributed decision support model to provide more intelligent support for IoT applications. The model is designed with three major approaches: an artificial neural network (ANN) for environment recognition, knowledge merging to create a local knowledge base and expert systems technology for decision making. In addition, a self-adaption feature is introduced to fix any possible improper usage of knowledge that may be caused by inaccuracies in the environment recognition. This strategy was confirmed in an experiment with a local Chinese medical clinical trials centre, in which the results indicate it may improve the accuracy of decisions from 42% to about 85%.