Scheduled Theoretical Restoration for Mining Immensely Partial Data Sets
Journal Title: The International Journal of Technological Exploration and Learning - Year 2013, Vol 2, Issue 5
Abstract
Partial data sets have turn out to be just about ubiquitous in an extensive range of application fields. Mutual illustrations can be initiate in climate and image data sets, sensor data sets, and medical data sets. The partiality in these data sets may stand up from a number of issues: In some circumstances, it may merely be a replication of definite measurements not being obtainable at the time, in others, the data may be absent due to incomplete system failure, or it may merely be a consequence of users being reluctant to stipulate attributes due to confidentiality worries. When a important portion of the entries are lost in all of the attributes, it turn into very tough to perform any generous of sensible extrapolation on the unique data. For such circumstances, we present the innovative idea of theoretical restoration in which we make effective theoretical representations on which the data mining algorithms can be openly smeared. The desirability behind the idea of theoretical restoration is to practice the correlation structure of the data in directive to precise it in terms of ideas rather than the unique dimensions. As a outcome, the restoration procedure evaluates only those theoretical aspects of the data can be mined from the partial data set, rather than might faults formed by extrapolation. We reveal the efficiency of the method on a range of actual data sets.
Authors and Affiliations
K. A. VarunKumar| Department of Computer Science and Engineering Vel Tech Dr.RR & Dr.SR Technical University, Chennai, S. Sibi Chakkaravarthy| Department of Computer Science and Engineering Vel Tech Dr.RR & Dr.SR Technical University, Chennai, M. Prabakaran| Department of Electrical & Electronics Engineering Vel Tech Dr.RR & Dr.SR Technical University, Chennai, Ajay Kaurav| Department of Electrical & Electronics Engineering Vel Tech Dr.RR & Dr.SR Technical University, Chennai, R. Baskar| Department of Electrical & Electronics Engineering Vel Tech Dr.RR & Dr.SR Technical University, Chennai, N. Nandhakishore| Department of Electrical & Electronics Engineering Vel Tech Dr.RR & Dr.SR Technical University, Chennai
Keywords
Related Articles
- EP ID EP8502
- DOI -
- Views 425
- Downloads 22
Data Aggregation Approach Using Neural Network in Wireless Sensor Networks
A wireless sensor network is one of the busiest networks because of multicast and broadcast network .In case of separate communication it gives high energy loss because there is requirement of some mechanism that can...
A Survey on: Utilizing of Different Features in Web Behavior Prediction
with the growing popularity of the World Wide Web, A large number of users uses web sites in the world. There are many technique which have been widely used to represent and analyze user‘s navigational behavior (usa...
Underwater Data Acquisition And Marine Boundary Indication System For Fishermen
Security and safety is the major area where many innovations are found to be developing in our life. Every nation wants to protect its border and its people. In order to achieve this, every country is taking serious s...
Fisher Algorithm: Variations And Applications
This paper examines Fisherface (Linear Discriminant Analysis) methods, its different modifications as applied to feature extraction in face recognition. Researchers showed that Fisher algorithm, though it performs be...
Enhanced Space Time Block Coded Spatial Modulation with Space-Time Trellis Codes For Sensing MIMO Technique
Multiple-input multiple-output systems have better performance under fading channels than single-input singleoutput (SISO) systems because of the huge capacity and reliability gains promised even in worst fading enviro...