Classical optimization software and consultancy - Classical optimization methods use a central node which has the entire responsibility or coordinating responsibility of deciding the optimal or near optimal solution to the problem.
|
Classical Optimization - Optimization
|
|
Classical DOE - Classical Design Of Experiments
|
|
Classical Design of Experiments shows you how to use Full Factorial experiments, Latin Square designs, Orthogonal Array designs as well as Sequential Experimentation. Each of these has unique strengths and weaknesses which can be used selectively for efficient research.
Classical Design of Experiments allows you to:
- Avoid a one-factor-at-a-time experiment
- Understand Latin Squares - square arrangements of factors
- Understand Plackett-Burman designs
- Understand Full Factorials - conducting experiments with factors and all their interactions
- One half fractional
- One quarter fractional
- One sixteenth fractional
- Exploit Sequential experimentation - building smaller experiments into larger experiments
- Dealiasing 1 main and all its 2-factor interactions
- Dealiasing all main effects
- Exploit Orthogonal Arrays - arrays that provide a mutually independent factor assignment
|
|
Classical DOE - Regression Analysis
|
|
Regression Analysis is a powerful technique of explaining relations between variables and how they affect a response. Advanced methods allow Multiple Regression Functions and Multiple Correlation analysis.
Regression Analysis allows you to:
- Perform Simple linear Regression analysis including
- linear Regression analysis
- Correlation analysis
- Residual analysis
- Lack of fit and pure error analysis
- Learn the matrix approach
- Perform Multiple Regression analysis including
- linear Coefficients analysis
- Sums of squares analysis
- Multiple Correlation analysis
|
|
|
Classical DOE - Stepwise Regression Analysis
|
|
Stepwise Regression Analysis is an extension to the Regression Analysis. This method is used when there are several independent variables (x1, x2, x3,...) and one dependent variable. Two methods are available: step-up and step-down.
Stepwise Regression Analysis allows you to:
- Perform a Step-up Regression
- Conduct a correlation analysis
- Select the highest correlated variable
- Fit the equation
- Iterate the procedure
- Perform a Step-down Regression
- Conduct a correlation analysis
- Select the highest correlated variable
- Fit the equation
- Iterate the procedure
|
|
|
|
Classical DOE - Orthogonal Polynomials Curve Fitting
|
|
Orthogonal Polynomial Curve Fitting is an important method of curve fitting for non linear functions. Non linear functions are usually expressed in some linear form using Taylor's theorem. Orthogonal Polynomial Curve Fitting shows how this may be done more efficiently by using orthogonal polynomial functions. Orthogonal Polynomial Curve Fitting allows you to:
- Fit Orthogonal Polynomials to
- Unifactor functions
- Multifactor functions
- Multifactor functions with interactions
- Learn why Taylor's Expansions are not as efficient
- Fit functions to complex cases
|
|
|