Features
New features in Global Optimization 10
Global Optimization 10 features three new regression functions and improved performance.
 Multimodel regression for comparing model fit to data in an AIC and adjusted goodnessoffit context
 Multistep regression to find the model with the best fit from a candidate set of terms
 Logistic regression
New features in Global Optimization 8
Global Optimization 8 has
enhanced performance and reliability, including improved parallel computing, module handling, and bug fixes.
 Utilizes parallel computing to speed up execution
 Solves problems when initial values or part of the search space are complex
 Accepts subscripted variable names
 General nonlinear constrained optimization solvers
 Accepts black box models (nonanalytic)
 Can solve problems in parallel mode
 Builtin nonlinear regression (with confidence intervals and sensitivity analysis)
 Maximum likelihood estimation
 Detailed user manual with examples
General features
 Handles constrained problems and problems with over 20,000
variables
 Solves problems with nonreal regions
 Solves constrained nonlinear regression problems using Chisquare, L1,
or L2 norms
 Solves maximumlikelihood statistical problems
 Solves very complex optimization problems
 Optimizes financial returns
 Solves enterprisecritical problems with high reliability
 Hillclimbing algorithms can solve nonlinear functions with
analytic equality and inequality constraints; can also solve
constrained (including boundconstrained) and unconstrained nonlinear
functions
 Solves problems using interval methods
 Solves 01 integer problems with a linear or nonlinear objective
function
 Solves smaller constrained or unconstrained global nonlinear
models
 A feasible starting point is not required in order to solve a
problem

