MultiscaleGaussianRFConfig

class franken.config.MultiscaleGaussianRFConfig(num_random_features, length_scale_low=4, length_scale_high=20, length_scale_num=6, use_offset=True, rng_seed=1337)

Bases: RFConfig

Random features to approximate a Gaussian kernel with multiple length-scales.

length_scale_high: float = 20

Top of the range for the length-scale. Should be positive.

length_scale_low: float = 4

Bottom of the range for the length-scale. Should be positive.

length_scale_num: int = 6

Number of different length-scales to include within the specified range.

num_random_features: int

Number of random features

rng_seed: int = 1337

Random seed for reproducibility of the random features

use_offset: bool = True

Whether or not to use the version of Gaussian random features with a uniform offset. Set to true, increases the estimator’s efficiency.