Transcript IBM Intelligent Miner introduction

IBM Intelligent Miner
introduction
Advisor: Dr. Hsu
Graduates: Yan-cheng Lin
Yu-Wei Su
2001/11/13
the lab of intelligent database
system, IDS
Outline










Main function
Data settings
Discretization settings
Mining settings
Name Mapping settings
Processing settings
Sequence settings
Statistics settings
Taxonomy settings
Value Mapping settings
2001/11/13
the lab of intelligent database
system, IDS
Main function









Data settings
Discretization settings
Mining settings
Name Mapping settings
Processing settings
Sequence settings
Statistics settings
Taxonomy settings
Value Mapping settings
2001/11/13
the lab of intelligent database
system, IDS
Data settings


Database table/view(DB2)
Flat files
2001/11/13
the lab of intelligent database
system, IDS
Discretization settings


To distribute records in the input data by
splitting value range of a continuous field into
intervals and the mapping each interval to a
discrete value
processing step




2001/11/13
Select discretization technique
Select input data if require
Specifying parameters for the discretization
Specifying the name of discretization
the lab of intelligent database
system, IDS
Mining settings






Associations
Classification
Clustering
Prediction
Sequential pattern
Similar sequences
2001/11/13
the lab of intelligent database
system, IDS
Name Mapping settings


Use name mappings to substitute value in
one field with the value in another field
Processing step



Selecting input data
Specifying parameters for the name mapping
Specifying the name of the name mapping
2001/11/13
the lab of intelligent database
system, IDS
Processing settings







Run SQL statement
Aggregate value
Encode missing/nonvalid value
Discard records with missing values
Filter fields/records
Get random sample
Join data sources
2001/11/13
the lab of intelligent database
system, IDS
Sequence settings





Mining function
Processing function
Sequence objects
Statistics function
Processing step



2001/11/13
Specifying the name of the sequence
Specifying the setting contained in the sequence
Specifying the parameters for this sequence
the lab of intelligent database
system, IDS
Statistics settings





bivariate statistic(二元變異統計)
principle component analysis(主分量分
析)
factor analysis(因子分析)
univariate curve(單變異曲線)
linear regression(線性迴歸)
2001/11/13
the lab of intelligent database
system, IDS
Taxonomy settings


It is a hierarchy of relationships
between categories
Processing step


2001/11/13
Specifying parameters for the taxonomy
Specifying the name for the taxonomy
the lab of intelligent database
system, IDS
Value Mapping settings


Use value mappings to substitute value
in one field with the value in another
field
Processing step



2001/11/13
Selecting input data
Specifying parameters for the value
mapping
Specifying the name of the value mapping
the lab of intelligent database
system, IDS
demonstration
2001/11/13
the lab of intelligent database
system, IDS