Skip to content

getml.feature_learning.Fastboost dataclass 

Fastboost(
    gamma: float = 0.0,
    loss_function: Optional[
        Union[CrossEntropyLossType, SquareLossType]
    ] = None,
    max_depth: int = 5,
    min_child_weights: float = 1.0,
    num_features: int = 100,
    num_threads: int = 1,
    reg_lambda: float = 1.0,
    seed: int = 5543,
    shrinkage: float = 0.1,
    silent: bool = True,
    subsample: float = 1.0,
)

Bases: _FeatureLearner

Feature learning based on Gradient Boosting.

Fastboost automates feature learning for relational data and time series. The algorithm used is slightly simpler than Relboost and much faster.

Enterprise edition

This feature is exclusive to the Enterprise edition and is not available in the Community edition. Discover the benefits of the Enterprise edition and compare their features.

For licensing information and technical support, please contact us.

PARAMETER DESCRIPTION
gamma

During the training of Fastboost, which is based on gradient tree boosting, this value serves as the minimum improvement in terms of the loss_function required for a split of the tree to be applied. Larger gamma will lead to fewer partitions of the tree and a more conservative algorithm. Range: [0, ∞]

TYPE: float DEFAULT: 0.0

loss_function

Objective function used by the feature learning algorithm to optimize your features. For regression problems use SquareLoss and for classification problems use CrossEntropyLoss.

TYPE: Optional[Union[CrossEntropyLossType, SquareLossType]] DEFAULT: None

max_depth

Maximum depth of the trees generated during the gradient tree boosting. Deeper trees will result in more complex models and increase the risk of overfitting. Range: [0, ∞]

TYPE: int DEFAULT: 5

min_child_weights

Determines the minimum sum of the weights a subcondition should apply to in order for it to be considered. Higher values lead to less complex statements and less danger of overfitting. Range: [1, ∞]

TYPE: float DEFAULT: 1.0

num_features

Number of features generated by the feature learning algorithm. Range: [1, ∞]

TYPE: int DEFAULT: 100

num_threads

Number of threads used by the feature learning algorithm. If set to zero or a negative value, the number of threads will be determined automatically by the getML Engine. Range: [0, ∞]

TYPE: int DEFAULT: 1

reg_lambda

L2 regularization on the weights in the gradient boosting routine. This is one of the most important hyperparameters in the Relboost as it allows for the most direct regularization. Larger values will make the resulting model more conservative. Range: [0, ∞]

TYPE: float DEFAULT: 1.0

seed

Seed used for the random number generator that underlies the sampling procedure to make the calculation reproducible. Internally, a seed of None will be mapped to 5543. Range: [0, ∞]

TYPE: int DEFAULT: 5543

shrinkage

Since Fastboost works using a gradient-boosting-like algorithm, shrinkage (or learning rate) scales down the weights and thus the impact of each new tree. This gives more room for future ones to improve the overall performance of the model in this greedy algorithm. It must be between 0.0 and 1.0 with higher values leading to a higher danger of overfitting. Range: [0, 1]

TYPE: float DEFAULT: 0.1

silent

Controls the logging during training.

TYPE: bool DEFAULT: True

subsample

Fastboost uses a bootstrapping procedure (sampling with replacement) to train each of the features. The sampling factor is proportional to the share of the samples randomly drawn from the population table every time Fastboost generates a new feature. A lower sampling factor (but still greater than 0.0), will lead to less danger of overfitting, less complex statements and faster training. When set to 1.0, the number of samples drawn will be identical to the size of the population table. When set to 0.0, there will be no sampling at all. Range: [0, 1]

TYPE: float DEFAULT: 1.0