LinearRegressionWithSGD

class pyspark.mllib.regression.LinearRegressionWithSGD

Train a linear regression model with no regularization using Stochastic Gradient Descent.

Methods

train(data[, iterations, step, …])

Train a linear regression model using Stochastic Gradient Descent (SGD).

Methods Documentation

classmethod train(data: pyspark.rdd.RDD[pyspark.mllib.regression.LabeledPoint], iterations: int = 100, step: float = 1.0, miniBatchFraction: float = 1.0, initialWeights: Optional[VectorLike] = None, regParam: float = 0.0, regType: Optional[str] = None, intercept: bool = False, validateData: bool = True, convergenceTol: float = 0.001) → pyspark.mllib.regression.LinearRegressionModel

Train a linear regression model using Stochastic Gradient Descent (SGD). This solves the least squares regression formulation

f(weights) = 1/(2n) ||A weights - y||^2

which is the mean squared error. Here the data matrix has n rows, and the input RDD holds the set of rows of A, each with its corresponding right hand side label y. See also the documentation for the precise formulation.

Parameters
datapyspark.RDD

The training data, an RDD of LabeledPoint.

iterationsint, optional

The number of iterations. (default: 100)

stepfloat, optional

The step parameter used in SGD. (default: 1.0)

miniBatchFractionfloat, optional

Fraction of data to be used for each SGD iteration. (default: 1.0)

initialWeightspyspark.mllib.linalg.Vector or convertible, optional

The initial weights. (default: None)

regParamfloat, optional

The regularizer parameter. (default: 0.0)

regTypestr, optional

The type of regularizer used for training our model. Supported values:

  • “l1” for using L1 regularization

  • “l2” for using L2 regularization

  • None for no regularization (default)

interceptbool, optional

Boolean parameter which indicates the use or not of the augmented representation for training data (i.e., whether bias features are activated or not). (default: False)

validateDatabool, optional

Boolean parameter which indicates if the algorithm should validate data before training. (default: True)

convergenceTolfloat, optional

A condition which decides iteration termination. (default: 0.001)

previous

LinearRegressionModel

next

RidgeRegressionModel