A Survey on Feature Selection Using FAST Approach to Reduce High Dimensional Data
Journal Title: INTERNATIONAL JOURNAL OF ENGINEERING TRENDS AND TECHNOLOGY - Year 2014, Vol 8, Issue 5
Abstract
Feature selection is a process of identifying the most useful subset of features.The survey summarising most of the feature selection methods and algorithms. Feature selection is the process of identifying a subset of most useful features. Typically the feature selection methods consist of four basic main steps and classify different existing feature selection algorithms .It is defined in terms of generation methods and evaluation functions. Most useful or representative methods are chosen from each category. The strength and weakness of different feature selection algorithms are explained. The aim here is to select some of the feature to form a feature subset. Feature selection has been effective technique in dimensionality reduction, removing irrelevant data, increasing learning accuracy, and improving comprehensibility. Increase in dimensionality of data imposes a severe challenge to many existing feature selection methods with respect to efficiency and effectiveness. To find a subset of features, the efficiency is related to time, the effectiveness is related to the quality of the subset of features. Existing feature selection algorithm removes only irrelevant features. But FAST algorithm removes both Irrelevant and redundant features. This survey mainly focuses on Comparison of various techniques and algorithms for feature selection process.
Authors and Affiliations
R. Munieswari , S. Saranya
Keywords
Related Articles
- EP ID EP162711
- DOI -
- Views 114
- Downloads 0
A comparative study of Mixed CNT bundle with Copper for VLSI Interconnect at 32nm
The aggressive technology scaling in VLSI leads to decrease the size of chip. Such continual miniaturization of VLSI devices has strong impact on the VLSI technology in several ways. The performance of ICs have b...
To Study and Analyze to foresee market Using Data Mining Technique
In every field there is huge growth and demand in knowledge and information over the internet. The automation using data mining and predictive technologies are doing an advance amount of deals in the markets. Data...
Experimental Study of Home Automation by Bicycle Pedal Power Using of Different Sprocket
This study attempts to measure the optimal pedaling rates for given power output levels as well as design the optimal number of gears and the corresponding gear ratios. With respect to human performance and power efficie...
Extraction of Core Contents from Web Pages
The information available on web pages mostly contains semi-structured text documents which are represented either in XML, or HTML, or XHTML format that lacks formatted document structure. The document does not discrimin...
Privacy and Utility in Data Publishing with Full Functional Dependencies
A number of methods have recently been proposed for privacy preserving data mining of multidimensional data records. One of the methods for privacy preserving data mining is that of anonymization, in which a reco...