Predicting software aging related bugs from imbalanced datasets by using data mining techniques
Journal Title: IOSR Journals (IOSR Journal of Computer Engineering) - Year 2016, Vol 18, Issue 1
Abstract
Abstract: Software aging bugs are related with the life-span of the software. Rebooting is one of the solutions of this problem, however, it is time consuming and causes resources loss. It is difficult to detect these bugsduring the time-limited software testing process. Data mining techniques can be useful to predict whether a piece of software has aging related bugs or not. The available datasets of software aging bugs present a challenge as they are imbalanced datasets. In these datasets, the number of data points with bugs is very small as compared to the number of data points with no bugs. It is important to predict the rare class (Bugs). In this paper we carried out experiment with a dataset containing data points related to aging-related bugs found in an open-source project MySQL DBMS. Data mining techniques developed for imbalanced datasets were compared with general data mining techniques. Various performance measures were used for the comparative study. The results suggest that data mining techniques developed for imbalanced datasets are more useful for correct prediction of data points related to aging related bugs. Data mining techniques developed for imbalanced datasets performed better than general data mining techniques on G-mean measure which is an important performance measure for imbalanced datasets
Authors and Affiliations
Amir Ahmad
Keywords
Related Articles
- EP ID EP143618
- DOI -
- Views 105
- Downloads 0
An Effective Approach to Discern Leaders in Wireless Sensor Networks
The influential significance across individual vertices in a Wireless Network raises a concern for the effective evaluation for Centrality metrics. A graph-theoretical comparison between the behaviour of betweenness, clo...
Prototyping and Simulation of Robot Group Intelligence using Kohonen Networks
Abstract:Intelligent agents such as robots can form ad hoc networks and replace human being in many dangerous scenarios such as a complicated disaster relief site. This project prototypes and builds a computer simulator...
User Profile Completion with Online Social Circles
User profile is a kind of most effective feature and cue of user behavioranalysis. However, most ofusers'profiles are incomplete and untruthful, and users' tag and profile data are rathersparse on social media. It...
Empirical Coding for Curvature Based Linear Representation inImage Retrieval System
Abstract : Image retrieval systems are finding their applications in all automation systems, wherein automateddecision needs to be taken based on the image contents. The prime requirement of such systems is to develop av...
Emblematical image based pattern recognition paradigm using Multi-Layer Perceptron Neural Network
Abstract: The abstract Likewise human brain machine can be signifying diverse pattern sculpt that is proficiently identify an image based object like optical character, hand character image, fingerprint and something lik...