This function needs to refer to the Sklearn package
Import Sklearn from Import Learning_curve
The calling format for this function is:
Learning_curve (estimator, X, Y, Train_sizes=array ([0.1 , 0.325, 0.55, 0.775, 1. ]), cv= None, Scoring=none, Exploit_incremental_learning=false, N_jobs=1, pre_dispatch='all' , Verbose=0)
The function is: for different sizes of training sets, determine the cross-validation training and test scores. A cross-validation generator divides the entire data set into K-times, dividing it into training sets and test sets. A subset of the different size training sets will be used to train the evaluator and a score will be generated for each training subset of the size, and then the score of the test set is calculated. Then, for each subset of training, all of the scores after running K will be averaged.
Estimator: the classifier used
X: Array-like, Shape (n_samples, N_features)
Training vectors, N_samples is the number of samples, N_features is the number of features
y: array-like, Shape (n_samples) or (N_samples, n_features), optional
Target relative to x classification or regression
train_sizes: Array-like, Shape (n_ticks,), dtype float or int
The relative or absolute numbers of the training samples, the samples of these quantities will generate learning curve. If Dtype is a float, he will be considered part of the maximum number of training sets (this is determined by the authentication method chosen). Otherwise, he will be considered the absolute size of the training set. It is important to note that for classification, the size of the sample must be large enough to contain at least one sample for each classification.
CV: int, cross-validation generator or an iterable, optional
Determine the separation strategy for cross-validation
--none, using the default 3-fold cross-validation,
--integer, make sure it's a few fold cross-validation
--an object that acts as a cross-validation generator
--an iterator that is applied to the training/test separation
Verbose:integer, optional
Control redundancy: The higher, the more information
return value:
train_sizes_abs: Array, shape = (n_unique_ticks,), Dtype int
The number of samples used to generate the training set for learning curve. Because duplicate input will be deleted, ticks may be less than n_ticks.
train_scores : Array, shape (n_ticks, N_cv_folds)
The score on the training set
test_scores : Array, shape (n_ticks, N_cv_folds)
The score on the test set
Python learning_curve function