I am just stepping into the research world after spending decades in industry - far away from academics. My apologies if my question is not a valid question or if the answers are already well known to experienced researchers.
Several methods/techniques are available for feature selection. Are there any recommended techniques or best practices to identify the most useful features?
I think you should try Random Forest approach.
You can find some useful information at http://cran.r-project.org/web/views/MachineLearning.html
I also advise you to have a look at caret package http://caret.r-forge.r-project.org/
You can also find interesting examples at http://www.r-bloggers.com/?s=feature+selection
Regards
My sincere and heartfelt thanks to all the members for your valuable contribution. After a few months gap, I am starting my project again. Thanks for your help/advice.
the terminology in this field can be similar and confusing. Feature selection (aka feature sub selection) selects an optimal sub-feature vector in order to reduce redundancy and co-linear behavior between features. Statistical tests such as mutual information between features is common and similarity thresholding. Feature selection is usually dividing into two categories: wrappers and filter.
Wrappers are driven by checking classification accuracy recursively as features are removed and tries many different combinations. Filters will filter information with tests such as mutual information, entropy, and feature co-linearity.
A similar method you may want to look at is feature extraction which involves transforming the original feature vector to a much smaller dimension, typically using PCA and/or LDA. This reduced set can then be sent to a classifier. Hope this helps.
Hi..
Feature Selection is basically selection of a subset of features from a larger pool of
available features. The goal is to select those that are rich in discriminatory information with respect to the classification problem at hand. This is a crucial step in the design of any classification system, as a poor choice of features drives the classifier to perform badly. Selecting highly informative features is an attempt
(a) to place classes in the feature space far apart from each other (large between-class distance) and
(b) to position the data points within each class close to each other (small within-class variance).
For more details: see chapter-4 of below book:
Book: An introduction to pattern recognition: A MATLAB approach
BY: Theodoridis, Sergios
You can also try WEKA data mining tool for feature selection, as in WEKA different feature selection algorithms are there.
Gud luck...!!!!
Dear Tamilalagan Natarajan,
I hope this link will be useful for you
Article A Feature Selection Model based on High-Performance Computin...
- Badges
- Science topic