Type KMeansClustering
Namespace tensorflow.contrib.factorization
Parent Estimator
Interfaces IKMeansClustering
An Estimator for K-Means clustering. Example:
```
import numpy as np
import tensorflow as tf num_points = 100
dimensions = 2
points = np.random.uniform(0, 1000, [num_points, dimensions]) def input_fn():
return tf.compat.v1.train.limit_epochs(
tf.convert_to_tensor(points, dtype=tf.float32), num_epochs=1) num_clusters = 5
kmeans = tf.contrib.factorization.KMeansClustering(
num_clusters=num_clusters, use_mini_batch=False) # train
num_iterations = 10
previous_centers = None
for _ in xrange(num_iterations):
kmeans.train(input_fn)
cluster_centers = kmeans.cluster_centers()
if previous_centers is not None:
print 'delta:', cluster_centers - previous_centers
previous_centers = cluster_centers
print 'score:', kmeans.score(input_fn)
print 'cluster centers:', cluster_centers # map the input points to their clusters
cluster_indices = list(kmeans.predict_cluster_index(input_fn))
for i, point in enumerate(points):
cluster_index = cluster_indices[i]
center = cluster_centers[cluster_index]
print 'point:', point, 'is in cluster', cluster_index, 'centered at', center
``` The `SavedModel` saved by the `export_savedmodel` method does not include the
cluster centers. However, the cluster centers may be retrieved by the
latest checkpoint saved during training. Specifically,
```
kmeans.cluster_centers()
```
is equivalent to
```
tf.train.load_variable(
kmeans.model_dir, KMeansClustering.CLUSTER_CENTERS_VAR_NAME)
```
Methods
Properties
- ALL_DISTANCES_dyn
- CLUSTER_CENTERS_VAR_NAME_dyn
- CLUSTER_INDEX_dyn
- config
- config_dyn
- COSINE_DISTANCE_dyn
- KMEANS_PLUS_PLUS_INIT_dyn
- model_dir
- model_dir_dyn
- model_fn
- model_fn_dyn
- params
- params_dyn
- PythonObject
- RANDOM_INIT_dyn
- SCORE_dyn
- SQUARED_EUCLIDEAN_DISTANCE_dyn
Fields
Public static methods
KMeansClustering NewDyn(object num_clusters, object model_dir, ImplicitContainer<T> initial_clusters, ImplicitContainer<T> distance_metric, ImplicitContainer<T> random_seed, ImplicitContainer<T> use_mini_batch, ImplicitContainer<T> mini_batch_steps_per_iteration, ImplicitContainer<T> kmeans_plus_plus_num_retries, object relative_tolerance, object config, object feature_columns)
Creates an Estimator for running KMeans training and inference. This Estimator implements the following variants of the K-means algorithm: If `use_mini_batch` is False, it runs standard full batch K-means. Each
training step runs a single iteration of K-Means and must process the full
input at once. To run in this mode, the `input_fn` passed to `train` must
return the entire input dataset. If `use_mini_batch` is True, it runs a generalization of the mini-batch
K-means algorithm. It runs multiple iterations, where each iteration is
composed of `mini_batch_steps_per_iteration` steps. Each training step
accumulates the contribution from one mini-batch into temporary storage.
Every `mini_batch_steps_per_iteration` steps, the cluster centers are
updated and the temporary storage cleared for the next iteration. Note
that:
* If `mini_batch_steps_per_iteration=1`, the algorithm reduces to the
standard K-means mini-batch algorithm.
* If `mini_batch_steps_per_iteration = num_inputs / batch_size`, the
algorithm becomes an asynchronous version of the full-batch algorithm.
However, there is no guarantee by this implementation that each input
is seen exactly once per iteration. Also, different updates are applied
asynchronously without locking. So this asynchronous version may not
behave exactly like a full-batch version.
Parameters
-
objectnum_clusters - An integer tensor specifying the number of clusters. This argument is ignored if `initial_clusters` is a tensor or numpy array.
-
objectmodel_dir - The directory to save the model results and log files.
-
ImplicitContainer<T>initial_clusters - Specifies how the initial cluster centers are chosen. One of the following: * a tensor or numpy array with the initial cluster centers. * a callable `f(inputs, k)` that selects and returns up to `k` centers from an input batch. `f` is free to return any number of centers from `0` to `k`. It will be invoked on successive input batches as necessary until all `num_clusters` centers are chosen. * `KMeansClustering.RANDOM_INIT`: Choose centers randomly from an input batch. If the batch size is less than `num_clusters` then the entire batch is chosen to be initial cluster centers and the remaining centers are chosen from successive input batches. * `KMeansClustering.KMEANS_PLUS_PLUS_INIT`: Use kmeans++ to choose centers from the first input batch. If the batch size is less than `num_clusters`, a TensorFlow runtime error occurs.
-
ImplicitContainer<T>distance_metric - The distance metric used for clustering. One of: * `KMeansClustering.SQUARED_EUCLIDEAN_DISTANCE`: Euclidean distance between vectors `u` and `v` is defined as \(||u - v||_2\) which is the square root of the sum of the absolute squares of the elements' difference. * `KMeansClustering.COSINE_DISTANCE`: Cosine distance between vectors `u` and `v` is defined as \(1 - (u. v) / (||u||_2 ||v||_2)\).
-
ImplicitContainer<T>random_seed - Python integer. Seed for PRNG used to initialize centers.
-
ImplicitContainer<T>use_mini_batch - A boolean specifying whether to use the mini-batch k-means algorithm. See explanation above.
-
ImplicitContainer<T>mini_batch_steps_per_iteration - The number of steps after which the updated cluster centers are synced back to a master copy. Used only if `use_mini_batch=True`. See explanation above.
-
ImplicitContainer<T>kmeans_plus_plus_num_retries - For each point that is sampled during kmeans++ initialization, this parameter specifies the number of additional points to draw from the current distribution before selecting the best. If a negative value is specified, a heuristic is used to sample `O(log(num_to_sample))` additional points. Used only if `initial_clusters=KMeansClustering.KMEANS_PLUS_PLUS_INIT`.
-
objectrelative_tolerance - A relative tolerance of change in the loss between iterations. Stops learning if the loss changes less than this amount. This may not work correctly if `use_mini_batch=True`.
-
objectconfig - See
tf.estimator.Estimator. -
objectfeature_columns - An optionable iterable containing all the feature columns used by the model. All items in the set should be feature column instances that can be passed to `tf.compat.v1.feature_column.input_layer`. If this is None, all features will be used.
Public properties
object ALL_DISTANCES_dyn get; set;
object CLUSTER_CENTERS_VAR_NAME_dyn get; set;
object CLUSTER_INDEX_dyn get; set;
object config get;
object config_dyn get;
object COSINE_DISTANCE_dyn get; set;
object KMEANS_PLUS_PLUS_INIT_dyn get; set;
object model_dir get;
object model_dir_dyn get;
object model_fn get;
object model_fn_dyn get;
object params get;
object params_dyn get;
object PythonObject get;
object RANDOM_INIT_dyn get; set;
object SCORE_dyn get; set;
object SQUARED_EUCLIDEAN_DISTANCE_dyn get; set;
Public fields
string SCORE
return string
|
string CLUSTER_INDEX
return string
|
string ALL_DISTANCES
return string
|