Mini-Batch K-Medoids clustering.
This method finds a set of cluster centers that are themselves data points, attempting to minimize the mean-squared distance from the datapoints to their assigned cluster centers using only mini-batches of the dataset.
Mini batches of the dataset are selected, and augmented to include each of the cluster centers. Then, standard KMedoids clustering is performed on the batch, using code based on the C clustering library [1]. The memory requirement scales as the square batch_size instead of the square of the size of the dataset.
| Parameters: | n_clusters : int, optional, default: 8
max_iter : int, optional, default=5
batch_size : int, optional, default: 100
metric : {“euclidean”, “sqeuclidean”, “cityblock”, “chebyshev”, “canberra”,
max_no_improvement : int, default: 10
random_state : integer or numpy.RandomState, optional
|
|---|
See also
References
| [R12] | de Hoon, Michiel JL, et al. “Open source clustering software.” Bioinformatics 20.9 (2004): 1453-1454. |
Attributes
| cluster_centers_ | (array, [n_clusters, n_features]) Coordinates of cluster centers |
| labels_ | (list of arrays, each of shape [sequence_length, ]) labels_[i] is an array of the labels of each point in sequence i. The label of each point is an integer in [0, n_clusters). |
Methods
| fit(sequences[, y]) | Fit the kcenters clustering on the data |
| fit_predict(sequences[, y]) | Performs clustering on X and returns cluster labels. |
| fit_transform(sequences[, y]) | Alias for fit_predict |
| get_params([deep]) | Get parameters for this estimator. |
| partial_predict(X[, y]) | Predict the closest cluster each sample in X belongs to. |
| partial_transform(X) | Alias for partial_predict |
| predict(sequences[, y]) | Predict the closest cluster each sample in each sequence in sequences belongs to. |
| set_params(**params) | Set the parameters of this estimator. |
| summarize() | |
| transform(sequences) | Alias for predict |
Fit the kcenters clustering on the data
| Parameters: | sequences : list of array-like, each of shape [sequence_length, n_features]
|
|---|---|
| Returns: | self |
Performs clustering on X and returns cluster labels.
| Parameters: | sequences : list of array-like, each of shape [sequence_length, n_features]
|
|---|---|
| Returns: | Y : list of ndarray, each of shape [sequence_length, ]
|
Alias for fit_predict
Get parameters for this estimator.
| Parameters: | deep: boolean, optional
|
|---|---|
| Returns: | params : mapping of string to any
|
Predict the closest cluster each sample in X belongs to.
In the vector quantization literature, cluster_centers_ is called the code book and each value returned by predict is the index of the closest code in the code book.
| Parameters: | X : array-like shape=(n_samples, n_features)
|
|---|---|
| Returns: | Y : array, shape=(n_samples,)
|
Alias for partial_predict
Predict the closest cluster each sample in each sequence in sequences belongs to.
In the vector quantization literature, cluster_centers_ is called the code book and each value returned by predict is the index of the closest code in the code book.
| Parameters: | sequences : list of array-like, each of shape [sequence_length, n_features]
|
|---|---|
| Returns: | Y : list of arrays, each of shape [sequence_length,]
|
Set the parameters of this estimator.
The method works on simple estimators as well as on nested objects (such as pipelines). The former have parameters of the form <component>__<parameter> so that it’s possible to update each component of a nested object.
| Returns: | self |
|---|
Alias for predict