Minimum Redundancy Feature Selection
- Author: O.Conrad (c) 2014
- Menu: Table | Calculus
Description
Identify the most relevant features for subsequent classification of tabular data.
The minimum Redundancy Maximum Relevance (mRMR) feature selection algorithm has been developed by Hanchuan Peng .
References:
Feature selection based on mutual information: criteria of max-dependency, max-relevance, and min-redundancy. Hanchuan Peng, Fuhui Long, and Chris Ding, IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 27, No. 8, pp.1226-1238, 2005.
Minimum redundancy feature selection from microarray gene expression data,
Chris Ding, and Hanchuan Peng, Journal of Bioinformatics and Computational Biology, Vol. 3, No. 2, pp.185-205, 2005.
Hanchuan Peng's mRMR Homepage at http://penglab.janelia.org/proj/mRMR/
Parameters
| | Name | Type | Identifier | Description | Constraints |
|---|
| Input | Data | table, input | DATA | - | - |
| Output | Feature Selection | table, output | SELECTION | - | - |
| Options | Class Identifier | table field | CLASS | - | - |
| Verbose Output | boolean | VERBOSE | output of intermediate results to execution message window | Default: 1 |
| Number of Features | integer number | mRMR_NFEATURES | - | Minimum: 1
Default: 50 |
| Discretization | boolean | mRMR_DISCRETIZE | uncheck this means no discretizaton (i.e. data is already integer) | Default: 1 |
| Discretization Threshold | floating point number | mRMR_THRESHOLD | a double number of the discretization threshold; set to 0 to make binarization | Minimum: 0.000000
Default: 1.000000 |
| Selection Method | choice | mRMR_METHOD | - | Available Choices:
[0] Mutual Information Difference (MID)
[1] Mutual Information Quotient (MIQ)
Default: 0 |
Command Line
Usage: saga_cmd table_calculus 12 [-DATA ] [-CLASS ] [-SELECTION ] [-VERBOSE ] [-mRMR_NFEATURES ] [-mRMR_DISCRETIZE ] [-mRMR_THRESHOLD ] [-mRMR_METHOD ]
-DATA: Data
table, input
-CLASS: Class Identifier
table field
-SELECTION: Feature Selection
table, output
-VERBOSE: Verbose Output
boolean
Default: 1
-mRMR_NFEATURES: Number of Features
integer number
Minimum: 1
Default: 50
-mRMR_DISCRETIZE: Discretization
boolean
Default: 1
-mRMR_THRESHOLD: Discretization Threshold
floating point number
Minimum: 0.000000
Default: 1.000000
-mRMR_METHOD: Selection Method
choice
Available Choices:
[0] Mutual Information Difference (MID)
[1] Mutual Information Quotient (MIQ)
Default: 0