Control systems architecture with a predictive identification model in digital ecosystems

Abstract

The paper describes basic architectural principals and main control system components using predictive identification models in digital ecosystems. We introduce the architecture for both Time-Driven and Batch-Driven and Alert-Driven modes for configuration of predictive identification models. In our work we discussed the main principals of Digital Ecosystems architecture with Alert-Driven control based on Associative search methods, regarding the main architectural components of each Ecosystem layer and its requirements for stability, reliability and scalability of such systems. In addition, the method of a predictive model development based on Data Mining approach with Associative Search is presented.