n_clusters : int, default 8, number of clusters

init : string, default "kmeans++", algorithm for initialization of clusters. *init should be "kmeans++" or "random".

n_jobs : int, default 1. the number of process when calculating k-means. *n_jobs > 1 is not implemented now.

maxiter : int, default 300, max iteration of Expectation and Maximization step of EM algorithm.

tol : float, default 0.001, convergence criterion.

X : array [n_samples, n_feature], data matrix used for trainig gmm. Each row corresponds to a single data point.

this : instance of GMM, which has fitted parameters.

n_components : int, default 1, the number of components of gmm

n_iter : int, default 100, max iteration of Expectation and Maximization step of EM algorithm.

thresh : float, default 0.01, convergence criterion.

min_covar : float, default 0.01, floor on the diagonal of the covariance matrix to prevent overfitting

X : array [n_samples, n_feature], data matrix used for training gmm. Each row corresponds to a single data point.

this : instance of GMM, which has fitted parameters.

X : array [n_samples, n_feature], data matrix.

loglikelihood : float, sum of negative log likelihood of given data X.

: None

: None

n_components : int, default 1, the number of dimension after reduction

whiten : boolean, default false, whether whitening is applied or not

X : array [n_samples, n_feature], data matrix to which dimension reduction is applied. Each row corresponds to a single data point.

this : instance of PCA, which has fitted parameters.

n_components : int, default 2, the number of dimension of latent space.

normalize : boolean, default true, whether data is centerized and normalized or not

X : array [n_samples, n_features], training samples of domain 1

Y : array [n_samples, n_features], training samples of domain 2

*the number of samples of X and Y should be the same.

this : instance of CCA, which has fitted parameters.

X : array [n_samples, n_features], test samples of domain 1

Y : array [n_samples, n_features], test samples of domain 2.

*the number of samples of X and Y should be the same.

X_score : array [n_samples, n_components], samples of domain1 after projection onto latent space.

Y_score : array [n_samples, n_components], samples of domain2 after projection onto latent space.

n_neightbors : int, default 5, the number of neighbors used to classify

algorithm : string, default "auto", algorithm of kNN

leaf_size : int, default 30, size of leaf

weights : string, default "uniform", weight of neighbors for prediction

*weights should be "uniform" or "dist".

X : array [n_samples, n_feature], training data matrix. each row corresponds to a single data point.

y : array [n_samples, 1], target vector. i-th element corresponds to the target value of i-th data of X.

this : instance of kNN Classifier, which has fitted parameters.

X : array [n_samples, n_feature], data matrix used for prediction.

res : array [n_samples, n_class], (i, j) element corresponds to predicted probability of i-th sample belonging j-th class.

center : boolean, default true, whether data is centerized or not

normalize : boolean, default true, whether data is normalized or not

solver : string, default 'qr', solver used to calculate parameter

*solver should be "lrqr" or "qr".

X : array [n_samples, n_feature], data matrix for training. each row corresponds to a single data point.

y : array [n_samples, 1], target vector. i-th element corresponds to the target value of i-th data of X.

this : instance of Ridge Regression, which has fitted parameters.

X : array [n_samples, n_feature], data matrix used for prediction.

pred : array [n_samples, 1], i-th element corresponds to the target value of i-th data of test data.

lambda : float, default 1.0, constant value to avoid overfitting

center : boolean, default true, whether data is centerized or not

normalize : boolean, default true, whether data is normalized or not

n_iter : int, default 1000, the number of iteration

tolerance : int, default 0.0001, convergence criterion of cooridate decent.

X : array [n_samples, n_feature], data matrix for training. each row corresponds to a single data point.

y : array [n_samples, 1], target vector. i-th element corresponds to the target value of i-th data of X.

this : instance of Ridge Regression, which has fitted parameters.

X : array [n_samples, n_feature], data matrix used for prediction.

pred : array [n_samples, 1], i-th element corresponds to the target value of i-th data of test data.

lambda : float, default 1.0, constant value for avoiding overfit

center : boolean, default true, whether data is centerized or not

normalize : boolean, default true, whether data is normalized or not

solver : string, default 'cd', solver used to calculate parameter

*solver should be "lrqr" or "cd" (coordinate decent).

n_iter : int, default 1000, the number of iterations

tolerance : int, default 0.0001, convergence criterion of cooridate decent.

X : array [n_samples, n_feature], data matrix for training. Each row corresponds to a single data point.

y : array [n_samples, 1], target vector. i-th element corresponds to the target value of i-th data of X.

this : instance of Ridge Regression, which has fitted parameters.

X : array [n_samples, n_feature], data matrix used for prediction.

pred : array [n_samples, 1], i-th element corresponds to the target value of i-th data of test data.

eta : float, default 0.01, learning rate for delta Error

alpha : float, default 0.015, l2-regularization strength

center : boolean, default true, whether data is centerized or not *center is not implemented.

n_iter : int, default 100, the number of iterations

X : array [n_samples, n_feature], data matrix for training. each row corresponds to a single data point.

y : array [n_samples, 1], target vector. i-th element corresponds to the target value of i-th data of X.

this : instance of Logistic Regression, which has fitted parameters.

X : array [n_samples, n_feature], data matrix used for prediction.

pred : array [n_samples, 1], i-th element corresponds to the target value of i-th data of test data.

algorithm : string, default 'sgdsvm', algorithm for online training

*algorithm should be "sgdsvm" or "perceptron".

n_iter : int, default 1000, the number of iterations

t_zero : float, default 1.0, value to decide step size alpha

aver : boolean, default true, whether averaging is used or not

X : array [n_samples, n_feature], data matrix of training samples. Each row corresponds to a single data point.

y : array [n_samples, n_class], label matrix of training samples. Each row corresponds to label information of single data.

init_w : array [n_feature+1, n_class], initial value of weight parameter

this : instance of Regressor, which has fitted parameters.

X : array [n_samples, n_feature], data matrix used for prediction.

pred : array [n_samples, 1], i-th element corresponds to the target value of i-th data of test data.

eta : float, default 1.0, learning rate

center : boolean, default true, whether data is centerized or not

n_iter : int, default 100, the number of iterations

X : array [n_samples, n_feature], data matrix of training samples. Each row corresponds to a single data point.

y : array [n_samples, 1], label vector of training samples. i-th element corresponds to the target value of i-th data of X.

this : instance of Perceptron, which has fitted parameters.

X : array [n_samples, n_feature], data matrix used for prediction.

pred : array [n_samples, 1], i-th element corresponds to the target value of i-th data of test data.

C : float, default 1.0, parameter which controls the influence of the regularization term

mode : string, default "PA2", mode of passive aggressive *mode should be "PA1" or "PA2".

n_iter : int, default 100, the number of iterations

X : array [n_samples, n_feature], data matrix of training samples. Each row corresponds to a single data point.

y : array [n_samples, 1], label vector of training samples. i-th element corresponds to the target value of i-th data of X.

this : instance of Perceptron, which has fitted parameters.

X : array [n_samples, n_feature], data matrix used for prediction.

pred : array [n_samples, 1], i-th element corresponds to the target value of i-th data of test data.

C : float, default 0.1, parameter which controls the influence of the regularization term

center : boolean, default true, whether data is centerized or not

n_iter : int, default 100, the number of iterations

X : array [n_samples, n_feature], data matrix of training samples. Each row corresponds to a single data point.

y : array [n_samples, 1], label vector of training samples. i-th element corresponds to the target value of i-th data of X.

this : instance of Perceptron, which has fitted parameters.

X : array [n_samples, n_feature], data matrix used for prediction.

pred : array [n_samples, 1], i-th element corresponds to the target value of i-th data of test data.

C : float, default 0.1, parameter which controls the influence of the regularization term

center : boolean, default true, whether data is centerized or not

n_iter : int, default 100, the number of iterations

X : array [n_samples, n_feature], data matrix of training samples. Each row corresponds to a single data point.

y : array [n_samples, 1], label vector of training samples. i-th element corresponds to the target value of i-th data of X.

this : instance of Perceptron, which has fitted parameters.

X : array [n_samples, n_feature], data matrix used for prediction.

pred : array [n_samples, 1], i-th element corresponds to the target value of i-th data of test data.

eta : float, default 0.9, probability that the sample is correctly classified

center : boolean, default true, whether data is centerized or not

n_iter : int, default 100, the number of iterations

X : array [n_samples, n_feature], data matrix of training samples. Each row corresponds to a single data point.

y : array [n_samples, 1], label vector of training samples. i-th element corresponds to the target value of i-th data of X.

this : instance of Perceptron, which has fitted parameters.

X : array [n_samples, n_feature], data matrix used for prediction.

pred : array [n_samples, 1], i-th element corresponds to the target value of i-th data of test data.

eta : float, default 0.9, probability that the sample is correctly classified

C : float, default 0.1, parameter which controls the influence of the regularization term

center : boolean, default true, whether data is centerized or not

n_iter : int, default 100, the number of iterations

X : array [n_samples, n_feature], data matrix of training samples. Each row corresponds to a single data point.

y : array [n_samples, 1], label vector of training samples. i-th element corresponds to the target value of i-th data of X.

this : instance of Perceptron, which has fitted parameters.

X : array [n_samples, n_feature], data matrix used for prediction.

pred : array [n_samples, 1], i-th element corresponds to the target value of i-th data of test data.