Academic Performance Evaluator over the Cluster-L2 Metric


Swati Jain1, Vikas Kumar Jain2, Sunil Kumar Kashyap3*, Sanjay Kumar1

1Department of Computer Science, Kalinga University, Raipur, Chhattisgarh, 492101, India

2Department of Chemistry, Government Engineering College, Raipur, Chhattisgarh, 492015, India

3Department of Mathematics, School of Advanced Sciences, Vellore Institute of Technology University, Vellore, Tamil Nadu, India, 632014

*Corresponding Author Email:



The Academic Performance of the Students interacts with the Data Management System. The performances of the student are recorded under some classes. These classes are well defined under some significant characteristics. There is various hidden information which lies with general set arrangements. The fuzzy set reforms this into membership function and fit into certain set theoretic representation. This paper proposes a new cluster analysis technique based on Euclidean Distance Metric (EDM) or L2 Metric. The advantages of this method are efficient and simple.






The set of information is called data. There may be big or small data. The big data can be characterized into the small data. If the big data is the set then the small data is referred as the subset or cluster. An analysis of these small data is called the cluster analysis. There are some methods of such analysis. Another new method is proposed in this paper. This method is based on EDM or Metric. Metric applied over the cluster. This cluster analyses by the statistical approach.


In 1971, Goodfellow [2] applied the similar type of technique to study the noncardiofarm bacteria. The result of this study was presented in the form of numerical taxonomy. Hartigan et al [3] studied the data of social changes via a new technique, which is referred later as k-means algorithm in the year 1979. Brossier [1] proposed another clustering method which is referred as piecewise hierarchical clustering in 1990. In 2002, Karypis [4] generated a toolkit for analysing the clusters referred as CLUTO. CLU for the clusters and To for the tools.



The data model is not just tabulation but this is an evaluator. The data is itself a self-mapped function. Its domain, co-domain and range are classified over the bilinear correspondence. Such mapping is adjoined and represented into the form of disjoint set which is countable and finite. This finiteness is lying with the physical output of information. Hence the data model is existed under some stochastic process. Definitely the classical metric has been providing several advantages but it is discrete. The proposed metric is continuous. The observation presented in the discrete form and the interpretation of the data will be presented in the continuous form.The proposed model is an arrangement of information which put the information in sequence and correspondence.



[1] Brossier G, Piecewise Hierarchical Clustering, Journal of classification, 7, 197-216, 1990.

[2] Goodfellow M, Numerical taxonomy of some nocardiofarm bacteria, Journal of general microbiology, 69(1), 33-80, 1971.

[3] Hartigan J, Wong M A, A K-means clustering algorithm, Applied Statistics, Royal Statistical Society, 28, 100-108, 1979.

[4] Karypis G, Cluto-A clustering toolkit, Technical Report 02017, University of Minnesota, Department of Computer Science, Minneapolis.






Received on 19.01.2017 Accepted on 18.02.2017

A&V Publications all right reserved

Research J. Engineering and Tech. 2017; 8(1): 01-03.

DOI: 10.5958/2321-581X.2017.00001.0