Creating a Regression Model

2 min read

Using the Transport Type option, you can select the target sediment load type.

With the DataList option, you can choose the desired ChildList for building the regression model. The data in the selected list will be displayed in the Selected DataList section.

The Input Unit option specifies the unit of the model’s input values, and the Output Unit option defines the unit for the model’s output values.

In the Equation Settings section, you can define the range for the two coefficients a and b.

These computations depend on two settings:

GA Settings (Genetic Algorithm Settings):
These settings configure the parameters for the Genetic Algorithm used to optimize the regression model. For more information, please refer to the section “Genetic Algorithm.”
Accuracy Settings:
These settings define the software’s criteria for evaluating and identifying better-performing models. To learn more, please refer to the section “Accuracy Evaluation Criteria.”

The Equation Type option specifies the type of sediment rating curve.

Two types of regression models or sediment rating curves can be fitted in this software:

Single-Line Rating Curves (USBR) #

it is possible to use the correction methods and coefficients FAO, QMLE, Smearing, Beta, and MVUE in this software.

By selecting the phrase Single-Line Rating Curve for the Equation Type option and clicking on the Run button, the genetic algorithm will attempt to find the two coefficients a and b, and will report the results to the user.

By clicking on the Enough button or by reaching the termination conditions defined in the settings, the computations will be completed, and the calculated results for the single-line rating curve will be displayed on the results page:

Multi-Line Rating Curves (S-curves) #

By selecting the phrase Multi-Line Rating Curve for the Equation Type option, the Number of Lines option will be activated, and the user will be able to select the number of curves desired for the model using this option. The Min Data Qty option will also be activated, allowing the user to specify the minimum number of data points required in each category.
By clicking on the Run button, the genetic algorithm will categorize the data into the number of curves selected using the Number of Lines option, and after finding the categorization with the largest SDR (Standard Deviation Reduction), it will report the results to the user.

By clicking on the Enough button or upon reaching the termination conditions defined in the calculation settings, the process will be completed, and the calculated results for the data categorization will be displayed on the following page:

By clicking on the Continue button, the genetic algorithm will attempt to find the two coefficients a and b for each curve and will report the results to the user.

By clicking on the Enough button or upon reaching the termination conditions defined in the calculation settings, the process will be completed, and the calculated results for the single-line rating curve will be displayed on the results page:

In the Results section on the results page, the outcomes and calculation details are reported, and in the Equation Parameters section, the coefficients and details of the equation are presented.

By right-clicking on the drawn graphs, the user can obtain a data report in Excel software and save the images using the Export Chart option.

For more information about this software page, please refer to the section “Export Charts.”

In the Description section, the user can write notes about the created model and save the created model in the project file using the Save This Equation option.