Genetic algorithm ga() was used to fit a model to data taken across 3D space.
In this problem, 6 parameters are fit for each spatial point, which takes up to 2 minutes on a laptop; there are 2000 points. Fitting the model to each and every point sequentially takes 1-2 days.
As a first-pass, each point was considered independently, even though the physics-based model parameters must vary smoothly over space. What is a computationally-efficient way to force this smooth spatial variation on all 6 parameters (ideally without requiring a supercomputer or several days of processing)?Genetic algorithm ga() was used to fit a model to data taken across 3D space.
In this problem, 6 parameters are fit for each spatial point, which takes up to 2 minutes on a laptop; there are 2000 points. Fitting the model to each and every point sequentially takes 1-2 days.
As a first-pass, each point was considered independently, even though the physics-based model parameters must vary smoothly over space. What is a computationally-efficient way to force this smooth spatial variation on all 6 parameters (ideally without requiring a supercomputer or several days of processing)? Genetic algorithm ga() was used to fit a model to data taken across 3D space.
In this problem, 6 parameters are fit for each spatial point, which takes up to 2 minutes on a laptop; there are 2000 points. Fitting the model to each and every point sequentially takes 1-2 days.
As a first-pass, each point was considered independently, even though the physics-based model parameters must vary smoothly over space. What is a computationally-efficient way to force this smooth spatial variation on all 6 parameters (ideally without requiring a supercomputer or several days of processing)? genetic algorithm, spatially varying coefficients, parameter estimation MATLAB Answers — New Questions

​