Title: Get Faster Results with Clementine 8'0
1Get Faster Results with Clementine 8.0
- Laura Squier
- Clementine Product Manager
- Thursday, October 16, 2003
2Agenda
- Clementine 8.0 themes
- New features for 8.0
- Demonstration
- Questions
3Themes for Clementine 8.0
- Enhancements for easier
- data preparation
- data and model understanding
- Enterprise level enhancements to support solution
based implementations
4Technology Goal
- Clementine is a scalable data mining workbench
that enables you to quickly develop predictive
models based on business expertise, and deploy
the models to improve decision making.
5Data Understanding
6Clementine 8.0 New Features
- New Data Audit Node
- The Data Audit node provides a comprehensive
first look at the data you bring into Clementine.
7Data Preparation
8Clementine 8.0 New Features
- New Field Reorder Node
- The Field Reorder node enables you to define the
natural order used to display fields downstream. - This order affects the display of fields in a
variety of places, such as tables, lists, and the
field chooser.
9Clementine 8.0 New Features
- Set random seed for sample
10Clementine 8.0 New Features
- New Binning Node
- The Binning node enables you to automatically
create new set fields based on the values of one
or more existing numeric range fields. - Mean and standard deviation
- Fixed-width binning
- Tiles (equal counts)
- Ranks
11Clementine 8.0 New Features
- New Reclassify Node
- The Reclassify node enables the transformation
from one set of discrete values to another.
Reclassification is useful for collapsing
categories or regrouping data for analysis.
12Clementine 8.0 New Features
- Enhanced Data Merging
- This release includes more sophisticated Merge
node capabilities. You can now merge records
using the following types of joins - Inner join
- Full outer join
- Partial outer join, both left and right joins
- Anti-join, the opposite of an inner join
13Modeling and Evaluation
14Clementine 8.0 New Features
- Lift Calculated for Apriori and GRI
15Clementine 8.0 New Features
- Generated Model Import
- Set up the stream in Clementine
- Export data to AnswerTree
- Access PMML and use in Clementine stream
16Clementine 8.0 New Features
- Improved Evaluation Charts
- Evaluation charts now include functionality
enabling you to define hit conditions and scoring
expressions used in the chart. You can also
specify a business rule condition used for
display.
17Clementine 8.0 New Features
- New Graphical Viewer for Cluster Models
- In this release, you can now view a graphical
representation of cluster results on the - Viewer tab for the following models
- Generated Kohonen net node
- Generated K-Means node
- Generated TwoStep Cluster
18Deployment
19Clementine 8.0 New Features
- New Deployment Options
- Deploy stream results into
- Cleo
- Predictive Marketing
- SMS
20Clementine 8.0 New Features
- Bulk Loading
- To increase performance during data export, you
can now use a custom bulk loader program specific
to your database. Options are available using the
Advanced button on Database and Publisher output
nodes.
21Clementine 8.0 New Features
- In-Database Scoring
- Clementine continues to expand its support for
in-database mining. In this release, you can now
generate SQL from decision trees and rulesets (in
addition to linear regression and factor
generated models). This allows scoring to be
conducted in the database, reducing costly data
transfer and calculations on the client.
22Integration
23Clementine 8.0 New Features
- CEMI Changes
- For those using the Clementine External Module
Interface (CEMI) to incorporate their own
functionality into Clementine, you can now create
and include custom CEMI icons for the node
palettes and generated model palettes.
24Clementine 8.0 New Features
- Improved SPSS and AnswerTree Integration
- Launch SPSS Server from Clementine Server
- Warn and correct field names
25Clementine 8.0 New Features
- Enhanced integration with Text Mining with
Clementine - Extract key concepts from documents for better
model performance
26Clementine 8.0 New Features
- Scripting and Batch Mode Changes
- Encoded Password Generator
- A tool is available through the user interface to
generate encoded passwords. Once encoded, you
can copy and store the password to script files
and command-line arguments. - Launch Using a Command File
- Command-line launch of Clementine and Clementine
Batch has been simplified with the use of the _at_
argument. To shorten or simplify the clemb or
clementine invocation command - Execute Script Selection
- You can now execute selected lines from a stream,
SuperNode, or standalone script using a new icon
on the toolbar.
27Demonstration
28Thank you
- Laura Squier
- Clementine Product Manager
- suggest_at_spss.com