Design of Experiment (DoE)
Planning and analyzing experiments to efficiently explore the effects of variables on outcomes.
Applications: Parameter screening, preparatory stages of sensitivity analysis, as well as model calibration and optimization.
Algorithms: Latin hypercube sampling, Full factorial design, Monte carlo sampling, Sobol' sequence, and more...
Problem & Benchmark
Defining user problems and provides benchmark cases to evaluate and compare algorithms.
Applications: Quick construction of user problems, Algorithm design and evaluation
Examples: Sphere, Rosenbrock, Ackley, ZDT kits, DTLZ kits, and more...
Surrogate Model
Approximating complex functions with simpler models to reduce computational cost and improve efficiency.
Applications: Reducing the computational cost of expensive simulations, Assisting analysis and optimization
Models: Polynomial regression, Gaussian process, Radial basis function, Kriging, Support vector regression, and more...
Analysis
Quantifying the specific influence of input uncertainties on model outputs.
Applications: Identifying key drivers of uncertainty, model simplification, risk assessment, and decision support.
Algorithms: Sobol', Fourier amplitude sensitivity test, Morris, Regional sensitivity analysis , Multivariate adaptive regression splines, and more...
Optimization
Searching for optimal designs or parameters, often coupled with surrogate modeling and sensitivity analysis.
Applications: Single objective optimization, multi objective optimization, surrogate-assisted optimization.
Algorithms: SCE-UA, Genetic algorithm, Differential evolution, Adaptive surrogate modelling-based optimization, NSGAII, MOEA/D, and more...
Inference
Estimating uncertain model parameters and quantifying posterior distributions.
Applications: Parameter estimation, Model calibration, Bayesian inference.
Algorithms: Markov chain monte carlo , Metropolis-Hastings , Metropolis-Hastings with Gibbs, Adaptive Metropolis-Hastings, DREAM-ZS, and more...
