Question 23
What is the purpose of the tol parameter of the SGDRegressor() in the given code below?
from sklearn.linear_model import SGDRegressormodel = SGDRegressor(early_stopping=True, validation_fraction=0.2, tol=0.001, n_iter_no_change=5)model.fit(X, y)It controls the learning rate of the stochastic regressor during training.
It determines the maximum number of iterations for the training process.
It defines the fraction of the validation set used for early stopping.
It specifies the tolerance level for early stopping based on the change in the validation error.