PCR Analyzer Software


Advanced Inputs Screen:

Optimization Parameters:

The program utilizes the Differential Evolution optimization routine promoted by Storn and Price (1995). For a detailed description of this routine and its parameters please refer to the referenced document.

Note: The default parameters should work well with most PCR data sets. You should not change these values unless you are having problems with the optimization routine.

NP
F
Cr
Range
Strategy


NP (Number of Population Vectors):

NP is the number of population vectors used in the optimization algorithm. In general the value should be 5 to 10 times the number of parameters. Sometimes even higher values are required.

Back to top


F (Weighting Factor):

F is the weighting factor used in the optimization algorithm. F is dependent on the Strategy choice, but in general the value should rarely be less than 0.5. Recommended values are 0.5 for the Best2Bin strategy and 0.7 to 0.8 for the Rand1Bin strategy.

Back to top


Cr (Crossover Constant):

The crossover constant Cr stems from the range (0,1) but is usually chosen to be either large (1.0) or small (0.2).

Back to top


Range:

The Range is the initial interval from which first set of parameter estimates are derived. Note this is a log-scale value.

Back to top


Strategy:

Strategy defines the optimization algorithm used. These choices are described in more detail in the referenced document.

Back to top

 

References:

Storn, Rainer and Price, Kenneth (1995). Differential Evolution - a Simple and Efficient Adaptive Scheme for Global Optimization over Continuous Spaces. Technical Report TR-95-012, ICSI, March 1995. PostScript-file downloadable from http://www.icsi.berkeley.edu/techreports/1995.abstracts/tr-95-012.html.