Oracle Data Mining functions and Algorithms

Oracle Data Mining functions and AlgorithmsOracle Data Mining API supports prediction and description mining functions. Prediction function, which uses training data to predict a target value. Describes a function to identify the inherent relationship between data. Each mining function specifies a type of problem to be solved. Each type can be implemented using one or more algorithms. API also provides basic data conversion tools to prepare data for mining. Oracle Data Mining prediction function

Function	Description	Example	Algorithm
Classification	A classification model that uses historical data to predict new discrete or classified data.	Provides a series of customer demographic statistics to predict the customer's response to the affinity card program.	Na has ve Bayes, adaptive Bayes Network, support vector machine, demo-tree
Anomaly Detection	Whether the irregular detection model predicts whether the data is in a typical location of a given distribution	A series of Customer statistics are given to identify different purchase behaviors between customers and normal customers.	PL/SQL and Java APIs currently support one type of SVM using classification mining functions and SVM algorithms without targets.
Regression	The decline model uses historical data to predict new continuous and digital data.	Provides a series of Customer statistics and purchase data to predict the customer's age	Support Vector Machine
Attribute importance	Attribute Value Model Recognition predicts the relationship between important attributes in a given result.	Gives the customer's response to the affinity card program and finds important independent attributes.	Minimal descriptor length

Oracle Data Mining description function

Function	Description	Example	Algorithm
Clustering	Natural grouping of clustered model recognition data sets	Segmented statistics are collected to 10 aggregates, and independent aggregation is learned to classify as much data as possible.	Enhanced K-means, orthogonal Clustering
Association Rules	Associate models to identify dataset relationships and their potential	Find the relationship between the items purchased by the customer	Apriori
Feature Extraction	Feature Extraction model creates an optimized dataset based on the model	A series of Customer statistics are given to extract typical features from the dataset.	Non-Negative Matrix Factorization