Fastboost

class getml.feature_learning.Fastboost(gamma: float = 0.0, loss_function: Optional[Literal['CrossEntropyLoss', 'SquareLoss']] = 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)

Feature learning based on Gradient Boosting.

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

Args:

gamma (float, optional):

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, \(\infty\)]

loss_function (loss_functions, optional):

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

max_depth (int, optional):

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, \(\infty\)]

min_child_weights (float, optional):

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, \(\infty\)]

num_features (int, optional):

Number of features generated by the feature learning algorithm. Range: [1, \(\infty\)]

num_threads (int, optional):

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: [\(0\), \(\infty\)]

reg_lambda (float, optional):

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, \(\infty\)]

seed (Union[int,None], optional):

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, \(\infty\)]

shrinkage (float, optional):

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]

silent (bool, optional):

Controls the logging during training.

subsample (float, optional):

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]

Note:

Not supported in the getML community edition.

Methods

validate([params])

Checks both the types and the values of all instance variables and raises an exception if something is off.

Attributes

gamma
loss_function
max_depth
min_child_weights
num_features
num_threads
reg_lambda
seed
shrinkage
silent
subsample
type