Advanced Settings

The Advanced Settings dialog is used to review and change the preference settings for out-of-sample data and genetic optimization.

 Access by clicking Advanced Settings from the Strategies tab of the Customize Studies & Strategies dialog.

Out-of-sample Window

Specifies the data that should not appear in the out-of-sample window. For the Walk-Forward Optimizer to function properly, you should exclude a portion of the data during optimization. It is suggested that you do not exclude less than 30% of the data and no more than 50% of the data during the Exhaustive/Genetic optimization.

  • Exclude first [ ] % and last [ ] % of data during optimization - Select this setting to exclude the first and/or last percentage of data, and enter the percentage amounts to exclude from optimization. When the optimization type is set to Standard, the default setting of 0% for both values will optimize on all data. When the type is set to Walk-Forward, it is recommended that you start with the first % set to 0 and the last % set to 30.
  • Exclude data before [ ] and after [ ] - Select this setting to exclude the before and after dates, and select the desired dates from the drop-down calendar.
  • Suggest - TradeStation will recommend in-sample and out-of-sample dates for the current chart range.

Genetic Optimization Settings

Contains the settings to customize the optimized strategy using the Genetic method. This will allow you to modify the genetic engine settings. Genetic Optimization Settings section is only enabled if you selected the Genetic method in the Optimization Details section of the Customize Studies & Strategies > Strategies tab dialog.

 Unless you have a good understanding of genetic algorithms, it is recommended that you only modify the Stress Test size and Stress Increment fields. Stress testing is a localized robust search around each individual parameter set within a genetic search, and only available when the Genetic Optimization method is used.

  • Generations - The number of iterations that the genetic algorithm will run until it arrives to the strongest population.
  • Population size - The number of chromosomes that will be subject to the genetic selection process. After the specified generations, the outstanding population group is understood as being of improved fitness due to the natural selection of the genetic algorithm.
  • Stress test size - The number (size) which defines the number of different parameter sets included with each stress test. If the Stress test size is one, then stress testing is turned off since no additional testing will be performed relative to the base parameter set. If you want to perform stress testing, the recommended setting is 3.
  • Mutation rate - The random rate of modification of the genes in chromosomes after each generation. Mutation is used to find new chromosomes to replace the 'weak' ones.
  • Crossover rate - The rate that specifies how close the replacement genes in chromosomes are picked after each generation. Crossover is used to find new chromosomes to replace the 'weak' ones.
  • Stress increment - The percentage that the parameters will be stressed either upwards or downwards.
  • Terminate Optimization if population fitness did not improve for [ ] generations - Check this setting for the genetic optimizer to automatically terminate if the population fitness did not improve for the specified number of generation. If no improvement took place for 10 (default) or more generations, the algorithm has converged and optimization is not necessary.

 Click OK to apply changes or Cancel to exit without changes.