Skip to main content
Oracle Help Center

Sign Out

Sign In

  • Table of Contents
  • Search
  • Print
  • Download
  1. Database
  2. Oracle
  3. Oracle Database
  4. Release 19

Concepts

  • Table of Contents
  • Search
  • Print
  • Download

Table of Contents

Expand AllCollapse All
  • Title and Copyright Information
  • Preface
    • Audience
    • Documentation Accessibility
    • Related Documentation
      • Oracle Data Mining Resources on the Oracle Technology Network
      • Application Development and Database Administration Documentation
    • Conventions
  • Changes in This Release for Oracle Data Mining Concepts Guide
    • Changes in Oracle Data Mining 18c
      • New Features
      • Deprecated Features
      • Desupported Features
      • Other Changes
  • Part I Introductions
    • 1 What Is Data Mining?
      • 1.1 What Is Data Mining?
        • 1.1.1 Automatic Discovery
        • 1.1.2 Prediction
        • 1.1.3 Grouping
        • 1.1.4 Actionable Information
        • 1.1.5 Data Mining and Statistics
        • 1.1.6 Data Mining and OLAP
        • 1.1.7 Data Mining and Data Warehousing
      • 1.2 What Can Data Mining Do and Not Do?
        • 1.2.1 Asking the Right Questions
        • 1.2.2 Understanding Your Data
      • 1.3 The Data Mining Process
        • 1.3.1 Problem Definition
        • 1.3.2 Data Gathering, Preparation, and Feature Engineering
        • 1.3.3 Model Building and Evaluation
        • 1.3.4 Knowledge Deployment
    • 2 Introduction to Oracle Data Mining
      • 2.1 About Oracle Data Mining
      • 2.2 Data Mining in the Database Kernel
      • 2.3 Data Mining in Oracle Exadata
      • 2.4 About Partitioned Model
      • 2.5 Interfaces to Oracle Data Mining
        • 2.5.1 PL/SQL API
        • 2.5.2 SQL Functions
        • 2.5.3 Oracle Data Miner
        • 2.5.4 Predictive Analytics
      • 2.6 Overview of Database Analytics
    • 3 Oracle Data Mining Basics
      • 3.1 Mining Functions
        • 3.1.1 Supervised Data Mining
          • 3.1.1.1 Supervised Learning: Testing
          • 3.1.1.2 Supervised Learning: Scoring
        • 3.1.2 Unsupervised Data Mining
          • 3.1.2.1 Unsupervised Learning: Scoring
      • 3.2 Algorithms
        • 3.2.1 Oracle Data Mining Supervised Algorithms
        • 3.2.2 Oracle Data Mining Unsupervised Algorithms
      • 3.3 Data Preparation
        • 3.3.1 Oracle Data Mining Simplifies Data Preparation
        • 3.3.2 Case Data
          • 3.3.2.1 Nested Data
        • 3.3.3 Text Data
      • 3.4 In-Database Scoring
        • 3.4.1 Parallel Execution and Ease of Administration
        • 3.4.2 SQL Functions for Model Apply and Dynamic Scoring
  • Part II Mining Functions
    • 4 Regression
      • 4.1 About Regression
        • 4.1.1 How Does Regression Work?
          • 4.1.1.1 Linear Regression
          • 4.1.1.2 Multivariate Linear Regression
          • 4.1.1.3 Regression Coefficients
          • 4.1.1.4 Nonlinear Regression
          • 4.1.1.5 Multivariate Nonlinear Regression
          • 4.1.1.6 Confidence Bounds
      • 4.2 Testing a Regression Model
        • 4.2.1 Regression Statistics
          • 4.2.1.1 Root Mean Squared Error
          • 4.2.1.2 Mean Absolute Error
      • 4.3 Regression Algorithms
    • 5 Classification
      • 5.1 About Classification
      • 5.2 Testing a Classification Model
        • 5.2.1 Confusion Matrix
        • 5.2.2 Lift
          • 5.2.2.1 Lift Statistics
        • 5.2.3 Receiver Operating Characteristic (ROC)
          • 5.2.3.1 The ROC Curve
          • 5.2.3.2 Area Under the Curve
          • 5.2.3.3 ROC and Model Bias
          • 5.2.3.4 ROC Statistics
      • 5.3 Biasing a Classification Model
        • 5.3.1 Costs
          • 5.3.1.1 Costs Versus Accuracy
          • 5.3.1.2 Positive and Negative Classes
          • 5.3.1.3 Assigning Costs and Benefits
        • 5.3.2 Priors and Class Weights
      • 5.4 Classification Algorithms
    • 6 Anomaly Detection
      • 6.1 About Anomaly Detection
        • 6.1.1 One-Class Classification
        • 6.1.2 Anomaly Detection for Single-Class Data
        • 6.1.3 Anomaly Detection for Finding Outliers
      • 6.2 Anomaly Detection Algorithm
    • 7 Clustering
      • 7.1 About Clustering
        • 7.1.1 How are Clusters Computed?
        • 7.1.2 Scoring New Data
        • 7.1.3 Hierarchical Clustering
          • 7.1.3.1 Rules
          • 7.1.3.2 Support and Confidence
      • 7.2 Evaluating a Clustering Model
      • 7.3 Clustering Algorithms
    • 8 Association
      • 8.1 About Association
        • 8.1.1 Association Rules
        • 8.1.2 Market-Basket Analysis
        • 8.1.3 Association Rules and eCommerce
      • 8.2 Transactional Data
      • 8.3 Association Algorithm
    • 9 Feature Selection and Extraction
      • 9.1 Finding the Best Attributes
      • 9.2 About Feature Selection and Attribute Importance
        • 9.2.1 Attribute Importance and Scoring
      • 9.3 About Feature Extraction
        • 9.3.1 Feature Extraction and Scoring
      • 9.4 Algorithms for Attribute Importance and Feature Extraction
    • 10 Time Series
      • 10.1 About Time Series
      • 10.2 Choosing a Time Series Model
      • 10.3 Time Series Statistics
        • 10.3.1 Conditional Log-Likelihood
        • 10.3.2 Mean Square Error (MSE) and Other Error Measures
        • 10.3.3 Irregular Time Series
        • 10.3.4 Build Apply
      • 10.4 Time Series Algorithm
  • Part III Algorithms
    • 11 Apriori
      • 11.1 About Apriori
      • 11.2 Association Rules and Frequent Itemsets
        • 11.2.1 Antecedent and Consequent
        • 11.2.2 Confidence
      • 11.3 Data Preparation for Apriori
        • 11.3.1 Native Transactional Data and Star Schemas
        • 11.3.2 Items and Collections
        • 11.3.3 Sparse Data
        • 11.3.4 Improved Sampling
          • 11.3.4.1 Sampling Implementation
      • 11.4 Calculating Association Rules
        • 11.4.1 Itemsets
        • 11.4.2 Frequent Itemsets
        • 11.4.3 Example: Calculating Rules from Frequent Itemsets
        • 11.4.4 Aggregates
        • 11.4.5 Example: Calculating Aggregates
        • 11.4.6 Including and Excluding Rules
        • 11.4.7 Performance Impact for Aggregates
      • 11.5 Evaluating Association Rules
        • 11.5.1 Support
        • 11.5.2 Minimum Support Count
        • 11.5.3 Confidence
        • 11.5.4 Reverse Confidence
        • 11.5.5 Lift
    • 12 CUR Matrix Decomposition
      • 12.1 About CUR Matrix Decomposition
      • 12.2 Singular Vectors
      • 12.3 Statistical Leverage Score
      • 12.4 Column (Attribute) Selection and Row Selection
      • 12.5 CUR Matrix Decomposition Algorithm Configuration
    • 13 Decision Tree
      • 13.1 About Decision Tree
        • 13.1.1 Decision Tree Rules
          • 13.1.1.1 Confidence and Support
        • 13.1.2 Advantages of Decision Trees
        • 13.1.3 XML for Decision Tree Models
      • 13.2 Growing a Decision Tree
        • 13.2.1 Splitting
        • 13.2.2 Cost Matrix
        • 13.2.3 Preventing Over-Fitting
      • 13.3 Tuning the Decision Tree Algorithm
      • 13.4 Data Preparation for Decision Tree
    • 14 Expectation Maximization
      • 14.1 About Expectation Maximization
        • 14.1.1 Expectation Step and Maximization Step
        • 14.1.2 Probability Density Estimation
      • 14.2 Algorithm Enhancements
        • 14.2.1 Scalability
        • 14.2.2 High Dimensionality
        • 14.2.3 Number of Components
        • 14.2.4 Parameter Initialization
        • 14.2.5 From Components to Clusters
      • 14.3 Configuring the Algorithm
      • 14.4 Data Preparation for Expectation Maximization
    • 15 Explicit Semantic Analysis
      • 15.1 About Explicit Semantic Analysis
        • 15.1.1 Scoring with ESA
        • 15.1.2 Scoring Large ESA Models
      • 15.2 ESA for Text Mining
      • 15.3 Data Preparation for ESA
      • 15.4 Terminologies in Explicit Semantic Analysis
    • 16 Exponential Smoothing
      • 16.1 About Exponential Smoothing
        • 16.1.1 Exponential Smoothing Models
        • 16.1.2 Simple Exponential Smoothing
        • 16.1.3 Models with Trend but No Seasonality
        • 16.1.4 Models with Seasonality but No Trend
        • 16.1.5 Models with Trend and Seasonality
        • 16.1.6 Prediction Intervals
      • 16.2 Data Preparation for Exponential Smoothing Models
        • 16.2.1 Input Data
        • 16.2.2 Accumulation
        • 16.2.3 Missing Value
        • 16.2.4 Prediction
        • 16.2.5 Parallellism by Partition
    • 17 Generalized Linear Models
      • 17.1 About Generalized Linear Models
      • 17.2 GLM in Oracle Data Mining
        • 17.2.1 Interpretability and Transparency
        • 17.2.2 Wide Data
        • 17.2.3 Confidence Bounds
        • 17.2.4 Ridge Regression
          • 17.2.4.1 Configuring Ridge Regression
          • 17.2.4.2 Ridge and Confidence Bounds
          • 17.2.4.3 Ridge and Data Preparation
      • 17.3 Scalable Feature Selection
        • 17.3.1 Feature Selection
          • 17.3.1.1 Configuring Feature Selection
          • 17.3.1.2 Feature Selection and Ridge Regression
        • 17.3.2 Feature Generation
          • 17.3.2.1 Configuring Feature Generation
      • 17.4 Tuning and Diagnostics for GLM
        • 17.4.1 Build Settings
        • 17.4.2 Diagnostics
          • 17.4.2.1 Coefficient Statistics
          • 17.4.2.2 Global Model Statistics
          • 17.4.2.3 Row Diagnostics
      • 17.5 GLM Solvers
      • 17.6 Data Preparation for GLM
        • 17.6.1 Data Preparation for Linear Regression
        • 17.6.2 Data Preparation for Logistic Regression
        • 17.6.3 Missing Values
      • 17.7 Linear Regression
        • 17.7.1 Coefficient Statistics for Linear Regression
        • 17.7.2 Global Model Statistics for Linear Regression
        • 17.7.3 Row Diagnostics for Linear Regression
      • 17.8 Logistic Regression
        • 17.8.1 Reference Class
        • 17.8.2 Class Weights
        • 17.8.3 Coefficient Statistics for Logistic Regression
        • 17.8.4 Global Model Statistics for Logistic Regression
        • 17.8.5 Row Diagnostics for Logistic Regression
    • 18 k-Means
      • 18.1 About k-Means
        • 18.1.1 Oracle Data Mining Enhanced k-Means
        • 18.1.2 Centroid
      • 18.2 k-Means Algorithm Configuration
      • 18.3 Data Preparation for k-Means
    • 19 Minimum Description Length
      • 19.1 About MDL
        • 19.1.1 Compression and Entropy
          • 19.1.1.1 Values of a Random Variable: Statistical Distribution
          • 19.1.1.2 Values of a Random Variable: Significant Predictors
          • 19.1.1.3 Total Entropy
        • 19.1.2 Model Size
        • 19.1.3 Model Selection
        • 19.1.4 The MDL Metric
      • 19.2 Data Preparation for MDL
    • 20 Naive Bayes
      • 20.1 About Naive Bayes
        • 20.1.1 Advantages of Naive Bayes
      • 20.2 Tuning a Naive Bayes Model
      • 20.3 Data Preparation for Naive Bayes
    • 21 Neural Network
      • 21.1 About Neural Network
        • 21.1.1 Neuron and activation function
        • 21.1.2 Loss or Cost function
        • 21.1.3 Forward-Backward Propagation
        • 21.1.4 Optimization Solver
        • 21.1.5 Regularization
        • 21.1.6 Convergence Check
        • 21.1.7 LBFGS_SCALE_HESSIAN
        • 21.1.8 NNET_HELDASIDE_MAX_FAIL
      • 21.2 Data Preparation for Neural Network
      • 21.3 Neural Network Algorithm Configuration
      • 21.4 Scoring with Neural Network
    • 22 Non-Negative Matrix Factorization
      • 22.1 About NMF
        • 22.1.1 Matrix Factorization
        • 22.1.2 Scoring with NMF
        • 22.1.3 Text Mining with NMF
      • 22.2 Tuning the NMF Algorithm
      • 22.3 Data Preparation for NMF
    • 23 O-Cluster
      • 23.1 About O-Cluster
        • 23.1.1 Partitioning Strategy
          • 23.1.1.1 Partitioning Numerical Attributes
          • 23.1.1.2 Partitioning Categorical Attributes
        • 23.1.2 Active Sampling
        • 23.1.3 Process Flow
        • 23.1.4 Scoring
      • 23.2 Tuning the O-Cluster Algorithm
      • 23.3 Data Preparation for O-Cluster
        • 23.3.1 User-Specified Data Preparation for O-Cluster
    • 24 R Extensibility
      • 24.1 Oracle Data Mining with R Extensibility
      • 24.2 Scoring with R
      • 24.3 About Algorithm Meta Data Registration
    • 25 Random Forest
      • 25.1 About Random Forest
      • 25.2 Building a Random Forest
      • 25.3 Scoring with Random Forest
    • 26 Singular Value Decomposition
      • 26.1 About Singular Value Decomposition
        • 26.1.1 Matrix Manipulation
        • 26.1.2 Low Rank Decomposition
        • 26.1.3 Scalability
      • 26.2 Configuring the Algorithm
        • 26.2.1 Model Size
        • 26.2.2 Performance
        • 26.2.3 PCA scoring
      • 26.3 Data Preparation for SVD
    • 27 Support Vector Machines
      • 27.1 About Support Vector Machines
        • 27.1.1 Advantages of SVM
        • 27.1.2 Advantages of SVM in Oracle Data Mining
          • 27.1.2.1 Usability
          • 27.1.2.2 Scalability
        • 27.1.3 Kernel-Based Learning
      • 27.2 Tuning an SVM Model
      • 27.3 Data Preparation for SVM
        • 27.3.1 Normalization
        • 27.3.2 SVM and Automatic Data Preparation
      • 27.4 SVM Classification
        • 27.4.1 Class Weights
      • 27.5 One-Class SVM
      • 27.6 SVM Regression
  • Glossary
  • Index

Search

Print

Download

PDF for offline and print

  • Previous
  • Next
  1. Concepts
  2. Introductions

Part I Introductions

Introduces Oracle Data Mining.

Provides a high-level overview for those who are new to data mining technology.

  • What Is Data Mining?

  • Introduction to Oracle Data Mining

  • Oracle Data Mining Basics

  • Previous
  • Next
Back to main content
  • About Oracle
  • Contact Us
  • Legal Notices
  • Terms of Use
  • Your Privacy Rights
  • Cookie Preferences
Copyright © 2005, 2019, Oracle and/or its affiliates. All rights reserved.
  • Previous
  • Next
Video