7 Clustering

This chapter describes clustering, the unsupervised mining function for discovering natural groupings in the data. For an overview of unsupervised data mining, see Chapter 3.

This chapter contains these topics:

About Clustering

Clustering analysis finds clusters of data objects that are similar in some sense to one another. The members of a cluster are more like each other than they are like members of other clusters. Different clusters can have members in common. The goal of clustering analysis is to find high-quality clusters such that the inter-cluster similarity is low and the intra-cluster similarity is high.

Clustering, like classification, is used to segment the data. Unlike classification, clustering models segment data into groups that were not previously defined. Classification models segment data by assigning it to previously-defined classes, which are specified in a target. Clustering models do not use a target.

Clustering is useful for exploring data. If there are many cases and no obvious groupings, clustering algorithms can be used to find natural groupings.

Clustering can serve as a useful data-preprocessing step to identify homogeneous groups on which to build supervised models.

Clustering can also be used for anomaly detection. Once the data has been segmented into clusters, you might find that some cases do not fit well into any clusters. These cases are anomalies or outliers.

How are Clusters Computed?

There are several different approaches to the computation of clusters. Oracle Data Mining supports the following methods:

  • Density-based — This type of clustering finds the underlying distribution of the data and estimates how areas of high density in the data correspond to peaks in the distribution. High-density areas are interpreted as clusters. Density-based cluster estimation is probabilistic.

  • Distance-based — This type of clustering uses a distance metric to determine similarity between data objects. The distance metric measures the distance between actual cases in the cluster and the prototypical case for the cluster. The prototypical case is known as the centroid.

  • Grid-based — This type of clustering divides the input space into hyper-rectangular cells and identifies adjacent high-density cells to form clusters.

Scoring New Data

Although clustering is an unsupervised mining function, Oracle Data Mining supports the scoring operation for clustering. New data is scored probabilistically.

Hierarchical Clustering

The clustering algorithms supported by Oracle Data Mining perform hierarchical clustering. The leaf clusters are the final clusters generated by the algorithm. Clusters higher up in the hierarchy are intermediate clusters.

Rules

Rules describe the data in each cluster. A rule is a conditional statement that captures the logic used to split a parent cluster into child clusters. A rule describes the conditions for a case to be assigned with some probability to a cluster.

Support and Confidence

Support and confidence are metrics that describe the relationships between clustering rules and cases. Support is the percentage of cases for which the rule holds. Confidence is the probability that a case described by this rule will actually be assigned to the cluster.

Evaluating a Clustering Model

Since known classes are not used in clustering, the interpretation of clusters can present difficulties. How do you know if the clusters can reliably be used for business decision making?

Oracle Data Mining clustering models support a high degree of model transparency. You can evaluate the model by examining information generated by the clustering algorithm: for example, the centroid of a distance-based cluster. Moreover, because the clustering process is hierarchical, you can evaluate the rules and other information related to each cluster's position in the hierarchy.

Clustering Algorithms

Oracle Data Mining supports these clustering algorithms:

The main characteristics of the two algorithms are compared in Table 7-1.

Table 7-1 Clustering Algorithms Compared

Feature k-Means O-Cluster Expectation Maximization

Clustering methodolgy

Distance-based

Grid-based

Distribution-based

Number of cases

Handles data sets of any size

More appropriate for data sets that have more than 500 cases. Handles large tables through active sampling

Handles data sets of any size

Number of attributes

More appropriate for data sets with a low number of attributes

More appropriate for data sets with a high number of attributes

Appropriate for data sets with many or few attributes

Number of clusters

User-specified

Automatically determined

Automatically determined

Hierarchical clustering

Yes

Yes

Yes

Probabilistic cluster assignment

Yes

Yes

Yes


Note:

Oracle Data Mining uses k-Means as the default clustering algorithm.