Customizing for Industrial Strength Applications

This chapter talks about the mathematical modeling of industrial processes. It looks at the hybrid combination of the mechanistic model and a data-driven neural network, as well as a look at variable driven processes, with practical examples from the industrial world.

Sections included in this chapter: Modeling, Combining Empirism with Mechanistic Understanding, Embedding an Idealized Model, Embedding a Priori Understanding, Incorporating Mixed Data Types, Confidence Measure for Characterizing Predictions, Interpreting Trained Neural Net Structures, and Graphical Interpretation.