Skip to Main Content
Sensor networks, peer-to-peer systems, and other large distributed systems, produce, store and process huge amounts of status data as a main part of their daily operation. Computing global predicates in such systems is usually very costly. The cost further increases when the data changes rapidly and computation has to follow these changes In this paper we describe a two step approach for dealing with these costs. First, we describe a highly efficient global algorithm which can be used to monitor a wide class of data mining models. Then, we use this algorithm as a feedback loop for the monitoring of complex functions of the data such as its outlier detection. Finally we would like to extend our future work is to show that how fail-stop process failures can be tolerated without check pointing or message logging.