A methodology for combining symbolic regression and design of experiments to improve empirical model building

摘要: A novel methodology for empirical model building using GPgenerated symbolic regression in combination with statistical design of experiments as well undesigned data is proposed. The main advantage this the maximum utilization when extrapolation necessary. offers alternative non-linear models that can either linearize response presence Lack or Fit challenge and confirm results from linear a cost effective time efficient fashion. economic benefit reduced number additional Fit.