Scikit-learn --- 5. Clustering Model

(A) General method parameters

1. General procedure

• get_params([deep]): Returns the parameters of the model.

  • deep: If Trueyou can return the child object model parameters.

• set_params(**params): Parameters of the model.

  • params: Keyword parameters to be set.

• fit(X[, y, sample_weight]) : Training model.

  • X: Sample set. Typically a numpy array, each row represents a sample and each column represents one feature.
  • y: Label sample collection. It and Xeach row corresponds to.
  • sample_weight: Weight of the sample weight. Its shape [n_samples,], each element represents a sample weight.

• predict(X, sample_weight): Returns the cluster label each sample belongs.

  • X: Sample set. Typically a numpy array, each row represents a sample and each column represents one feature.
  • sample_weight: Weight of the sample weight. Its shape [n_samples,], each element represents a sample weight.

• fit_predict(X[, y, sample_weight]) : Training model and implement cluster, returns cluster label each sample belongs.

  • X: Sample set. Typically a numpy array, each row represents a sample and each column represents one feature.
  • y: Label sample collection. It and Xeach row corresponds to.
  • sample_weight: Weight of the sample weight. Its shape [n_samples,], each element represents a sample weight.

• transform(X): The data set  X switch to cluster center space .

  In cluster center space dimensions, the sample is its distance from the center of each cluster.

  • X: Sample set. Typically a numpy array, each row represents a sample and each column represents one feature.

• fit_transform(X[, y, sample_weight]): Training model and implementation of clustering, the data set  X switch to cluster center space .

  • X: Sample set. Typically a numpy array, each row represents a sample and each column represents one feature.
  • y: Label sample collection. It and Xeach row corresponds to.
  • sample_weight: Weight of the sample weight. Its shape [n_samples,], each element represents a sample weight.

2.通用参数

• n_jobs: A positive number, and when the form specified tasks specified  CPUnumber.

  If  -1 you use all available  CPU.

• verbose: A positive number. For opening / closing the intermediate iteration output log function.

  • The larger the value, the more detailed the log.
  • Value of 0 or Nonedisables the log output.

• max_iter : An integer specifying the maximum number of iterations.

  If Nonecompared to the default value (different solverdifferent default value).

• tol: A floating-point number, specify a threshold algorithm converges.

• random_state: An integer or a RandomStateinstance, or None.

  • If an integer, it specifies the random number seed generator.
  • If as RandomStatean example, it specifies the random number generator.
  • If it is None, then the default random number generator.

A, KMeans

Two, DBSCAN

Three, MeanShift

Four, AgglomerativeClustering

Five, BIRCH

六、GaussianMixture

七、SpectralClustering