Indian Journal of Science and Technology

Open Access Open Access  Restricted Access Subscription Access

A Stage-by-Stage Pruning Method for Classifying Uncertain Data Streams


Affiliations
1 Department of Computer Science Engineering, Fatima Michael College of Engineering and Technology, Madurai – 625020, Tamil Nadu, India
2 Department of Information Technology, K.L.N. College of Information and Technology, Madurai – 630612, Tamil Nadu, India
 

Background: We study an important problem of similarity grouping processing on stream data that inherently contain uncertainty. Method: In this paper SBSP - [Stage by Stage Pruning] a novel pruning method is proposed for fast, accurate clustering and classifying the data where the two stages were grouped into a single framework MYFRAME. Findings: The proposed approach group the data-by-data level pruning using Manhattan distance in first stage. In the second stage, the data is grouped by object level pruning in hyperspace. Improvements: Currently, this approach is applied in real time applications such as object detection, video retrieval, people detection and tracking, earth quake monitoring etc.

Keywords

Clustering, Data Pruning, Distance, Group Nearest Neighbor, Grouping Process, Similarity Search, Uncertain Data Streams
User

Abstract Views: 208

PDF Views: 0




  • A Stage-by-Stage Pruning Method for Classifying Uncertain Data Streams

Abstract Views: 208  |  PDF Views: 0

Authors

S. Subashini
Department of Computer Science Engineering, Fatima Michael College of Engineering and Technology, Madurai – 625020, Tamil Nadu, India
S. Appavu alias Balamurugan
Department of Information Technology, K.L.N. College of Information and Technology, Madurai – 630612, Tamil Nadu, India

Abstract


Background: We study an important problem of similarity grouping processing on stream data that inherently contain uncertainty. Method: In this paper SBSP - [Stage by Stage Pruning] a novel pruning method is proposed for fast, accurate clustering and classifying the data where the two stages were grouped into a single framework MYFRAME. Findings: The proposed approach group the data-by-data level pruning using Manhattan distance in first stage. In the second stage, the data is grouped by object level pruning in hyperspace. Improvements: Currently, this approach is applied in real time applications such as object detection, video retrieval, people detection and tracking, earth quake monitoring etc.

Keywords


Clustering, Data Pruning, Distance, Group Nearest Neighbor, Grouping Process, Similarity Search, Uncertain Data Streams



DOI: https://doi.org/10.17485/ijst%2F2016%2Fv9i8%2F131021