1 Introduction
2 Motivation
3 Methodology
4 Problem statement
5 Related works
6 Objectives
7 Proposed system
7.1 Clustering data in the peer-to-peer environment using the hybrid map reduce and greedy-based local approximation fuzzy clustering approach
7.1.1 Clustering the selected data
7.1.2 Functions of K-means
-
Input: centroids input.//The input is centroid of the cluster and (value of the input) data points.
-
Output: The output is the nearest cluster of object and value of the object.
-
1: nxtCentroid ⇓ null, nxtDist ⇓ ∞
-
2: for each c ϵ Centroids do
-
3: dist ⋄ Distance (input. Value, c);
-
4: if nxtCentroid == null || dist < nxtDist then
-
5: nxtCentroid ⇓ c, nxtDist ⇓ dist;
-
6: end if
-
7: end for
-
8: output. Collect (nxtCentroid, object)
7.2 Estimating the accuracy of cluster using harmonic search
8 Performance analysis
8.1 Skin segmentation dataset
8.2 Adult dataset
Dataset characteristics | Univariate |
---|---|
Attribute characteristics | Real |
Associated tasks | Classification |
Number of instances | 245,057 |
Number of attributes | 4 |
Missing values | N/A |
Area | Computer |
Number of web hits | 106,563 |
8.3 Performance metrics
8.3.1 Jaccard index
8.3.2 F-measure
8.3.3 Mutual information
8.3.4 Rand measure
Data cluster | Skin segmentation | Adult database | ||
---|---|---|---|---|
Methods | Jaccard measure | Rand measure | Jaccard measure | Rand measure |
K-means | 0.67 | 0.54 | 0.73 | 0.62 |
MEKPFCM | 0.87 | 0.84 | 0.85 | 0.81 |
Proposed method | 0.93 | 0.91 | 0.95 | 0.93 |
Data cluster | Skin segmentation | Adult database | ||
---|---|---|---|---|
Method | Precision | Recall | Precision | Recall |
K-means | 0.75 | 0.61 | 0.77 | 0.65 |
MEKPFCM | 0.87 | 0.82 | 0.89 | 0.84 |
Proposed method | 0.94 | 0.92 | 0.96 | 0.91 |
Method | Skin segmentation | Adult dataset |
---|---|---|
K-means | 0.45 | 0.42 |
MEKPFCM | 0.21 | 0.19 |
Proposed method | 0.07 | 0.04 |
Method | Skin segmentation | Adult dataset |
---|---|---|
K-means | 79 | 81 |
MEKPFCM | 89 | 92 |
Proposed method | 96 | 97 |
Methods | Skin segmentation | Adult dataset |
---|---|---|
K-means | 3.8 | 2.9 |
MEKPFCM | 2.4 | 1.7 |
Proposed method | 1.4 | 1.0 |