Vendor Implementations - Java Data Mining: Strategy, Standard, and Practice - page 411

16.1.3 Oracle JDM Capabilities

OJDM supports all the JDM 1.1 mining functions: classification,
regression, attribute importance, association, and clustering . As an
Oracle extension, OJDM also supports feature extraction and one-
class classifier , also known as the mining function anomaly detection .
In OJDM, applications can optionally select algorithms and their
settings. Table 16-2 shows the OJDM supported functions, algo-
rithms, and tasks. OJDM also supports export and import tasks for
exporting and importing mining models in the Oracle Database
format. OJDM allows data mining users to specify function level
settings and have the database use a default algorithm.

Table 16-2 Oracle Java Data Mining supported functions, algorithms,

and function-level tasks

Supported function Supported algorithms Supported function level tasks

Classification Naïve Bayes (NB) BuildTask

Support Vector Machine (SVM) ClassificationTestTask

Decision Tree ClassificationTestMetricsTask

Adaptive Bayes Network (ABN) * DataSetApplyTask

RecordApplyTask

Regression Support Vector Machine (SVM) BuildTask

RegressionTestTask
RegressionTestMetricsTask
DataSetApplyTask
RecordApplyTask

Minimum Description Length *

Attribute importance BuildTask

Apriori *

Association BuildTask

Clustering k-Means BuildTask

Orthogonal Partitioning (a.k.a. O-Cluster) *

DataSetApplyTask
RecordApplyTask

Feature extraction Non-Negative Matrix Factorization

(NMF) *

BuildTask
DataSetApplyTask
RecordApplyTask

One-class classifier
(a.k.a. Anomaly
detection)

Support Vector Machine BuildTask

DataSetApplyTask
RecordApplyTask

Oracle-specific functions and algorithms