Adaptive model of data predictability in designing of information systems

The paper describes a model for a predictability analysis of various data types, which may be used to manage different real information systems. The models are “adaptive” in a sense of an absence of the fundamental assumptions about the probability distributions of the data. We present a special software tool based on the method of moving separation of the finite probability mixtures to ascertain the possibility of forecasting for the data.