Using AdvancedMiner

Table of Contents

1. Workflow
Introducing Workflow
Node descriptions
Connection status of nodes
Data sources
Data analysis
Technical Transformations (operations on data)
Analytical Transformations (Data Transformation)
2. Gython – the AdvancedMiner Scripting Language
Python quick reference
Flow control
Working with objects
Defining and calling functions
Gython methods for different types of variables
String methods
List methods
Dictionary methods
Python Library functions
Built-in functions
String functions
Mathematical functions
Random functions
Date/time objects
Managing Gython objects
Constructing and accessing objects
Saving objects
Loading objects
Renaming objects
Executing tasks
Deleting objects
Checking object existence
Task termination
Saving script environment
Loading script environment
Setting alias to the metadata repository
Sending messages to the log
Registry Repository
Project path
Context Scripts
How do context scripts work?
Where can I find context scripts?
Writing context scripts
Requesting user input using InputDialog
3. AdvancedMiner in Practice
Model building
General rules
Approximation model building
Classification model building
Clustering model building
Survival model building
Model testing
Approximation Test Task
Classification Test Task
Survival Test Task
Time Series Test Task
Classification Test Result Task
Applying Models in AdvancedMiner
Basic concepts
Advanced concepts
Minimal set-up
Applying for different mining functions
Shorthand methods of building, testing and applying models
Experiments project
Running experiments
Comparing models
Social Network Analysis
Building networks
Filtering networks
Analysing networks
Visualising networks
Building models by ABM
4. Data Access and Data Processing
Database Access
Database explorer
Using SQL statements
Importing and exporting data and other database operations
Importing MS Excel spreadsheets
Importing CSV files
Exporting data to MS Excel spreadsheets
Exporting data to CSV files
Getting column list for a database table
Deleting a database table
Checking for a database table existence.
Creating tables in Gython
Creating a table with manually specified data
Creating a table with data copied from a list
Creating a table with data obtained from an sql query
Using lists to define column names and formats
Importing data from external sources
The Trans procedure
Basic transformations
The where keyword
The keep in and drop in keywords
The keep out and drop out keywords
The format keyword
Flow control
Appending tables
The rename keyword
Joining tables
Data transformation functions
Ranking data (the rank procedure)
Expansion of data (the interpolate procedure)
Sampling data (the sample command)
Splitting tables (the tableSplit procedure)
Transposing tables (the transpose procedure)
Comparing two tables (the tablesCompare procedure)
Predefined transformations for data Mining models
Transformation Types
Important notes
5. Integration with common office suites
Built-in support for Office Suites
Setting up an MS-Office connection
Setting up an OpenOffice connection
Creating custom reports
Creating and working with a spreadsheet document
Creating and using a text document
6. Optimization Library
The Optimization Problem
Objective Function
Optimization methods
Solving the optimization problem
7. Statistical Procedures and Tests
Statistical functions
Chi-square statistic
Pearson's correlation coefficient
Multidimensional frequency analysis procedure
Statistical tests
Statistical test usage
Empirical distribution function
The Anderson-Darling test
The Chi-square test
The F-test
Kolmogorov-Smirnov test
Kuiper test
Levene's test
The Mann-Whitney test
Pearson's test
Test of proportions
Sign test
Spearman's test
Student's t-test
8. Probability distributions
Distributions Library
Characteristics and samples of the distributions
List of available continuous distributions
List of available discrete distributions
Distribution Tables
Special Functions Library
Sample Statistics of Empirical Data
Random Number Generators
9. Monte Carlo Markov Chains Library
The MarkovChain class
MarkovChain object methods
MarkovChain static methods
The Metropolis algorithm
Metropolis-Hastings algorithm
Bayesian inference
Transition functions
Transition functions from distribution
Random walk transition function
Sampling Distribution
Likelihood function
Helper distributions
Convergence Diagnostics and Output Analysis tool
Output Analysis
10. Scoring Code in AdvancedMiner
Scoring code for models
Creating Java scoring code based on a model step by step
Architecture of Java scoring code
Executing scoring code for a model
Differences in Scoring Code output for various model
Executing scoring code outside the AdvancedMiner system
Reading the Input Signature
Example of using scoring code in an external application
11. Data Visualization
Preparing data for plotting
Data objects
Declaring column types
Automatically obtaining the data type
Data specification patterns
Series grouping
Inconsistent data
Creating plots and charts
Chart objects
Chart object methods
Chart types
Grouping charts
Additional topics
Manipulating plots
Manipulating 2D plots
Manipulating 3D plots
12. Freq - a visual data exploration tool
Introducing Freq
Launching Freq
Overview of the Freq component
Working with attributes
Calculating attributes
Attribute view
Attribute display modes
Histogram types
Editing levels and grouping values
Analyzing data with Freq
Virtual attributes
Filtering data
Working with targets
Correlation matrix
Exporting to Excel spreadsheets
Attribute statistics in Freq
Basic attribute statistics
Attribute correlation statistics
Target related statistics
Integration with other components
Opening physical data
Viewing data
Binding between components
13. Report Engine
14. Operating Server
Requirements and Architecture
Quick Start guide
15. Model Reports
Efficiency Report
Statistical Test Report
Stability Report