#include <CscSupervisedLearningRegression.h>
|
| static float | findMaxLocal (const vector< float > &coeffPolynomial, float minBounding, float maxBounding, float precision) |
| | This method is used to find the local maximum of a polynomial function of one variable within a given interval. More...
|
| |
◆ CscSupervisedLearningRegression()
| CscSupervisedLearningRegression::CscSupervisedLearningRegression |
( |
RegressionInputsData |
regressionInputsData | ) |
|
◆ bestDegreePolynomialRegression()
This method iteratively trains polynomial regression models of increasing degree until the coefficient of determination (R^2) stops improving.
The method starts with a polynomial regression of degree 1 and continues to increase the degree until the R^2 value of the current model is less than the R^2 value of the previous model. The model with the highest R^2 value is then considered to be the best.
Note: This method assumes that the polynomialRegression method is implemented and works as expected.
- Returns
- The best trained model's output data as an object of the RegressionOutputsData class, containing the highest coefficient of determination and the corresponding coefficients of the polynomial function.
◆ findMaxLocal()
| float CscSupervisedLearningRegression::findMaxLocal |
( |
const vector< float > & |
coeffPolynomial, |
|
|
float |
minBounding, |
|
|
float |
maxBounding, |
|
|
float |
precision |
|
) |
| |
|
static |
This method is used to find the local maximum of a polynomial function of one variable within a given interval.
The polynomial function is defined by a vector of coefficients passed as an argument. The method evaluates the polynomial function at a series of points in the interval defined by minBounding and maxBounding, using the precision value to determine the spacing between these points. The method returns the x-value where the polynomial function reaches its local maximum within the given interval.
- Parameters
-
| coeffPolynomial | A vector of coefficients defining the polynomial function. First value correspond to coefficient of x⁰, second to x¹, third to x², etc. |
| minBounding | The lower bound of the interval within which to search for the local maximum. |
| maxBounding | The upper bound of the interval within which to search for the local maximum. |
| precision | The spacing between points in the interval at which the polynomial function is evaluated. A smaller value will result in a more precise maximum, but will increase computation time. |
- Returns
- The x-value within the given interval where the polynomial function reaches its local maximum.
◆ polynomialRegression()
This method performs polynomial regression on the given input data.
The method first transforms the given input data into the Eigen::MatrixXf format, then applies polynomial transformation on the input data up to the given degree. A linear regression model is then trained on this transformed data. The trained model is then tested on the input data and the coefficient of determination is calculated. If the dataToCsv flag is set, the method will print the results in a csv file.
- Parameters
-
| dataInputs | The input data for the regression in the form of regressionInputsData object. This object contains a vector of vectors, where each inner vector represents a data point and the elements of the inner vector represent the features of that point. |
| degree | The degree of the polynomial transformation to be applied on the input data. |
| dataToCsv | A boolean flag that, if set to true, will print the results in a csv file. |
- Returns
- A CscSupervisedLearningRegression::regressionOutputsData object that contains the coefficient of determination and the coefficients of the polynomial function obtained from the regression.
The documentation for this class was generated from the following files:
1.8.17