Autonomous Predictors, Estimators, Filters, Inferential Sensors
Plamen Parvanov Angelov · 2012
The autonomous learning systems (ALS) concept is quite generic and can be applied to numerous problems. This chapter describes one of the problem, namely, predictors, estimators, filters and inferential sensors. A specific of the regression models is that the input and output represent different physical variables. The role of ALS is to represent the nonlinear regression, especially in case of a nonstationary data stream. Another popular problem (apart from regression) is the time series. This problem is also widely used in forecasting, statistical learning and signal processing. Soft/intelligent/inferential sensors are widely used in industry due to their ability to provide accurate real-time estimates or predictions of difficult to measure variables of interest and replaces expensive measurements. An important part of the autonomous learning sensors design and development is the automatic input variables selection. Controlled Vocabulary Terms filters; learning systems; sensors; time series