Oracle قاعدة البيانات 19c: Data Mining Techniques LVC

• I Introduction
• Course Goals I-2
• Course Agenda I-3
• Summary I-4

Data Mining Concepts and Terminology

• Objectives
• What is data mining?
• Who uses data mining?
• Defining Key Data Mining Properties
• Data Mining Categories
• Examining Supervised Data Mining
• Supervised Data Mining Steps
• Supervised Data Mining Techniques
• Supervised Data Mining Algorithms
• Examining Unsupervised Data Mining
• Unsupervised Data Mining Techniques
• Unsupervised Data Mining Algorithms
• What Data Mining Can and Cannot Do
• Summary

The Data Mining Process

• Objectives
• Steps in the Data Mining Process
• Step 1: Defining the Problem
• Step 2: Selecting, Preparing, and Preprocessing Data
• Data Preparation Tips
• Data Preparation: Aggregation
• Data Preparation: Mining Attribute Types
• Automatic Data Preparation (ADP)
• Step 3: Building and Evaluating Models
• Model-Building Tasks
• Training and Testing Models: Supervised Learning
• Evaluating Model Results
• Applying the Selected Models
• Step 4: Deploying Results
• Summary

Introducing Oracle Data Miner

• Objectives
• Oracle Data Mining: Architecture
• Oracle Data Mining: Related Uses
• Oracle Data Miner
• Setting Up Oracle Data Miner (Using SQL Developer)
• Installing SQL Developer
• Setting Up Data Miner
• The Data Miner Repository Installation Process
• Creating a Data Miner Project and Workflow
• Introducing the Data Miner Interface
• Examining Oracle Data Miner Nodes
• Previewing a Data Miner Workflow
• Summary

Using Classification Models

• Objectives
• Revisiting Supervised Learning
• Supervised Learning: Examining Classification Techniques
• Using Classification Models
• Building a Workflow
• The Starting Point for Any Workflow: A Data Source
• Adding a Data Source to the Workflow
• Defining a Data Source: Step 1
• Viewing Source Data
• Defining a Data Source: Step 2
• Completed Data Source Node
• Previewing the Source Data
• Adding an Explore Data Node
• Connecting Workflow Nodes
• Specifying a “Group By” for the Explore Data Node
• Running the Explore Data Node
• Viewing the Explore Data Results
• Examining the Results
• Creating Classification Models
• Adding a Classification Node
• Classification Node Default Properties
• Connecting Data Source and Class Build Nodes
• Selecting Target and Case ID Attributes
• Viewing Advanced Settings
• Accessing Help
• Specifying Advanced Settings
• Specifying Test Properties
• Building the Models
• Successful Build Indicators
• Viewing the Event Log
• Comparing the Models
• Comparing Test Results: Performance
• Comparing Test Results: Lift
• Comparing Test Results: Performance Matrix
• Comparing Test Results: Profit
• Tuning a Model
• Selecting a Specific Model
• Navigating the Decision Tree Model Interface
• Examining the Decision Tree Test Results
• Viewing the Leaf Node in the Decision Tree
• Scoring: Applying the Model to Your Data
• Step 1: Specifying the Desired Model
• Step 2: Adding a New Data Source Node
• Modifying a Node Name
• Step 3: Adding an Apply Node
• Step 4: Connecting the Nodes
• Step 5: Creating the Output Format
• Step 6: Running the Apply Node
• Viewing Predictive Results
• Sorting Results
• Examining Predictions
• Summary

Using Regression Models

• Objectives
• Supervised Learning: Examining Regression Techniques
• Using Regression Models
• Creating a New Workflow
• Adding the Data Source
• Understanding Target and Input Variables
• Adding, Connecting, and Running an Explore Data Node
• Exploring Source Data Statistics
• Identifying Non-Input Variables
• Filtering the Source Data
• Deselecting Columns
• Running the Source Data Filter
• Creating Regression Models
• Adding a Regression Node
• Regression Node Default Properties
• Connecting and Editing the Regression Node
• Editing Advanced Settings: GLM Regression
• Viewing Advanced Settings: SVM Regression
• Building the Models
• Successful Build Indicators
• Viewing Models and Results
• Viewing GLM Model Details
• Viewing GLM Model Settings
• Viewing GLM Model Coefficients
• Viewing SVM Model Information
• Sorting Model Coefficients
• Viewing Test Results for the GLM Model
• GLM Model Test Results: Residual Tab
• GLM Model Test Results: Predicted and Actual
• Viewing Test Results for the SVM Model
• SVM Model Test Results: Predicted and Actual
• Comparing Test Results
• Comparing GLM and SVM Test Results
• Scoring Data with the Selected Model
• Step 1: Specifying the Chosen Model
• Step 2: Adding a New Data Source Node
• Removing Non-Input and Target Variables
• Modifying the Node Name
• Step 3: Adding an Apply Node
• Step 4: Connecting the Nodes
• Step 5a: Modifying the Output Format
• Step 5b: Modifying the Output Format
• Step 5c: Modifying the Output Format
• Step 6: Running the Apply Node
• Viewing Predictive Results
• Sorting Results
• Examining Predictions
• Summary

Using Clustering Models

• Objectives
• Revisiting Unsupervised Data Mining
• Unsupervised Data Mining: Clustering
• Using a Clustering Model
• Creating the New Workflow
• Using Existing Workflow Objects
• Connecting and Editing the Explore Data Node
• Specifying a “Group By” for the Explore Data Node
• Running and Viewing the Explore Data Node
• Exploring Source Data Statistics
• Creating Clustering Models
• Adding a Clustering Node
• Default Properties of Clustering Nodes
• Connecting the Nodes and Modifying Node Properties
• Examining Cluster Model Settings
• K-Means Advanced Algorithm Settings
• O-Cluster Advanced Algorithm Settings
• Expectation Maximization Advanced Algorithm Settings
• Building the Models
• Viewing the Models
• Viewing K-Means Model Results
• Navigating the K-Means Tree
• Viewing Results with Cluster Rule
• Viewing Other Cluster Definitions
• Viewing Cluster Details
• Comparing Clusters
• Multiple Row Selection with Cluster Compare
• Renaming Clusters
• Viewing the O-Cluster Model Results
• Viewing the Expectation Maximization Model Results
• Applying a Clustering Model
• Creating an Output Format: Apply Columns
• Creating an Output Format: Data Columns
• Running the Model and Viewing Data
• Viewing Results
• Viewing Sorted Results
• Summary

Performing Market Basket Analysis

• Objectives
• Unsupervised Data Mining: Association Rules
• Using an Association Rules Model
• Creating a New Project and Workflow
• Adding a Data Source from a New Schema
• Selecting a New Table or View
• Previewing the Data Source
• Examining Sorted Data
• Creating an Association Rules Model
• Adding an Association Node
• Connecting Data Source and Assoc Build Nodes
• Specifying Required Input Items
• Specifying Advanced Settings
• Building the Model
• Viewing the Model
• Viewing Model Results
• Examining Model Rules
• Creating Model Rule Filters
• Applying Model Rule Filters
• Viewing Model Itemsets
• Creating Model Details Output
• Viewing Model Details
• Using Descriptive Columns in the Output
• Adding a Join Node
• Defining the Join
• Specifying Output Columns for the Join
• Defining a New Assoc Build Node
• Viewing Model Results
• Viewing Descriptive Model Rules
• Creating New Model Details Output
• Viewing Descriptive Model Details
• Summary

Performing Anomaly Detection

• Objectives
• Unsupervised Data Mining: Anomaly Detection
• Using an Anomaly Detection Model
• Creating the New Workflow
• Adding a Data Source
• Browsing Source Data Attributes
• Previewing the Source Data
• Creating an Anomaly Detection Model
• Adding an Anomaly Detection Node
• Connecting the Nodes and Selecting the Case ID
• Specifying the SVM Algorithm Kernel Function
• Building the Model
• Viewing Model Results
• Viewing “Normal” Results
• Viewing “Anomalous” Results
• Comparing Model Results
• Applying the Model
• Creating an Output Format: Apply Columns
• Creating an Output Format: Data Columns
• Running the Model and Viewing Data
• Sorting Output Results
• Examining Results
• Summary

Deploying Data Mining Results

• Objectives
• Agenda
• Creating a Table or View of Results
• Modifying the Output Format
• Generating the Output and Viewing Results
• Viewing Output Results
• Using Output Results
• Agenda
• SQL Code Automatically Generated by Data Miner
• Deploying SQL Code to the Clipboard
• Accessing Deployed Code
• Executing Deployed Code
• Modifying Deployed Code
• Summary