Calibration Model Maintenance
Model maintenance can be defined as the on-going servicing of (primarily) multivariate calibration and fault detection models in order to preserve their predictive abilities. It is required because of changes to either the sample matrices or the instrument response. The goal of model maintenance is to sustain or improve models over time and changing conditions with the least amount of cost and effort. This course presents a roadmap for determining when model maintenance is required, the probable source of the response variations, and the appropriate approaches for achieving it. Hands-on exercises will be done using MATLAB and PLS_Toolbox.
1.0 Introduction to Model Maintenance 1.1 Goals 1.2 Causes of data/model mismatch 2.0 Identifying the need for action 2.1 External validation samples 2.2 Model prediction diagnostics Q and T^2 2.3 Setting action limits 2.4 Determining the source of the problem 3.0 Model updating methods 3.1 Slope and bias adjustments 3.2 Adding to the calibration set 3.3 Automatic model updating 4.0 Instrument standardization methods 4.1 Direct and Piecewise Direct standardization 4.2 Spectral Subspace Transformation 4.3 Filtering approaches: GLS and OSC 5.0 Avoiding model updating 5.1 Preprocessing methods that tend to make model more brittle 5.2 Using the model robustness tests 6.0 Conclusions