Conservative Noise Filters
Journal Title: International Journal of Advanced Computer Science & Applications - Year 2016, Vol 7, Issue 5
Abstract
Noisy training data have a huge negative impact on machine learning algorithms. Noise-filtering algorithms have been proposed to eliminate such noisy instances. In this work, we empirically show that the most popular noise-filtering algorithms have a large False Positive (FP) error rate. In other words, these noise filters mistakenly identify genuine instances as outliers and eliminate them. Therefore, we propose more conservative outlier identification criteria that improve the FP error rate and, thus, the performance of the noise filters. With the new filter, an instance is eliminated if and only if it is misclassified by a mutual decision of Naïve Bayesian (NB) classifier and the original filtering criteria being used. The number of genuine instances that are incorrectly eliminated is reduced as a result, thereby improving the classification accuracy.
Authors and Affiliations
Mona M. Jamjoom, Khalil El Hindi
Keywords
Related Articles
- EP ID EP101638
- DOI 10.14569/IJACSA.2016.070548
- Views 93
- Downloads 0
Database Preservation: The DBPreserve Approach
In many institutions relational databases are used as a tool for managing information related to day to day activities. Institutions may be required to keep the information stored in relational databases accessible becau...
Stable Haptic Rendering For Physics Engines Using Inter-Process Communication and Remote Virtual Coupling
Availability of physics engines has significantly reduced the effort required to develop interactive applications concerning the simulation of physical world. However, it becomes a problem when kinesthetic feedback is ne...
Analysis of Coauthorship Network in Political Science using Centrality Measures
In recent era, networks of data are growing massively and forming a shape of complex structure. Data scientists try to analyze different complex networks and utilize these networks to understand the complex structure of...
Test Case Generation For Concurrent Object-Oriented Systems Using Combinational Uml Models
Software testing is an important phase of software development to ensure the quality and reliability of the software. Due to some limitations of code based testing method, the researcher has been taken a new method...
Integrating Social Network Services with Vehicle Tracking Technologies
This paper gives design, and implementation of a newly proposed vehicle tracking system, that uses the popular social network as a value added service for traditional tracking system. The proposed tracking system make us...