The paper is devoted to the issues of data mining and building on their basis predictive models for systems of diagnostics and monitoring of the technological process of facilities of critical information infrastructure. The paper considers a method to information analysis for the construction of instant models of forecasting parameters of technological process with consideration, detailing coefficients of multiple-scale wavelet transformation. An example of the use of the proposed approach for prediction and assessment of the risk potential of the temperature of the feed water behind the high-pressure heater in the power unit of the NPP is presented.