The Importance of Feature Selection in Classification
Journal Title: International Journal on Computer Science and Engineering - Year 2014, Vol 6, Issue 1
Abstract
Feature Selection is an important technique for classification for reducing the dimensionality of feature space and it removes redundant, irrelevant, or noisy data. In this paper the feature are selected based on the ranking methods. (1) Information Gain (IG) attribute evaluation, (2) Gain Ratio (GR) attribute evaluation, (3) Symmetrical Uncertainty (SU) attribute evaluation. This paper evaluates the features which are derived from the 3 methods using supervised learning algorithms K-Nearest Neighbor and Naïve Bayes. The measures used for the classifier are True Positive, False Positive, Accuracy and they compared between the algorithm for experimental results. we have taken 2 data sets Pima and Wine from UCI Repository database.
Authors and Affiliations
Mrs. K. Moni Sushma Deep , Mr. P. Srinivasu
Keywords
Related Articles
- EP ID EP115549
- DOI -
- Views 151
- Downloads 0
Using Association Rule Mining for Extracting Product Sales Patterns in Retail Store Transactions
Computers and software play an integral part in the working of businesses and organisations. An immense amount of data is generated with the use of software. These large datasets need to be analysed for useful informatio...
An Analysis of Particle Swarm Optimization with Data Clustering-Technique for Optimization in Data Mining.
Data clustering is a popular approach for automatically finding classes, concepts, or groups of patterns. Clustering aims at representing large datasets by a fewer number of prototypes or clusters. It brings simplicity i...
IMAGE SEGMENTATION BY USING EDGE DETECTION
In this paper, we present methods for edge segmentation of atellite images; we used seven techniques for this category; obel operator technique, Prewitt technique, Kiresh technique, Laplacian technique, Canny technique...
Honeypot based Secure Network System
A honeypot is a non-production system, design to interact with cyber-attackers to collect intelligence on attack techniques and behaviors. There has been great amount of work done in the field of network intrusion detect...
Method for forecasting the volume of applications using probabilistic classifier based on Bayesian theorem for recruitment in the government organizations
In this paper, we propose a practical method to estimate or forecast the volume of applications based on Bayesian classifier which is a probability method that makes optimal decision based on known probability distributi...