Generates a labeled dataset whose samples form a hyperplane in n-dimensional vector space and whose labels are continuous values drawn from a uniform random distribution between -1 and 1. When the number of coefficients is either 1, 2 or 3, the samples form points, lines, and planes respectively. Due to its linearity, Hyperplane is especially useful for testing linear regression models.
Data Types: Continuous
Label Type: Continuous
|1||coefficients||[1, -1]||array||The n coefficients of the hyperplane where n is the dimensionality.|
|2||intercept||0.0||float||The y intercept term.|
|3||noise||0.1||float||The factor of gaussian noise to add to the data points.|
This generator does not have any additional methods.
use Rubix\ML\Datasets\Generators\Hyperplane; $generator = new Hyperplane([0.1, 3, -5, 0.01], 150.0, 0.25);