Scribd
Upload
58 views2 pages

Improved K-Mean Clustering With Steepest Ascent

The document proposes an improved K-means clustering algorithm using a steepest ascent (gradient) technique. It uses color descriptors in RGB and HSV color spaces as input. The improved K-means algorithm incorporates a heuristic local search method to provide additional information and find a better centroid value, resulting in more stable and effective clustering compared to conventional K-means. The proposed system iteratively assigns data points to the nearest centroid, recalculates the centroids, and uses an error function to evaluate clustering.

Uploaded by

Michael Benilan
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOC, PDF, TXT or read online on Scribd
58 views2 pages

Improved K-Mean Clustering With Steepest Ascent

The document proposes an improved K-means clustering algorithm using a steepest ascent (gradient) technique. It uses color descriptors in RGB and HSV color spaces as input. The improved K-means algorithm incorporates a heuristic local search method to provide additional information and find a better centroid value, resulting in more stable and effective clustering compared to conventional K-means. The proposed system iteratively assigns data points to the nearest centroid, recalculates the centroids, and uses an error function to evaluate clustering.

Uploaded by

Michael Benilan
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOC, PDF, TXT or read online on Scribd
You are on page 1/ 2

Improved K-Mean Clustering with Steepest Ascent (Gradient) Method for Image Retrieval

ABSTRACT:
In this study we present a new frame work for clustering that uses an Improved K-Mean with Steepest ascent (Gradient) Technique. The basic idea of this paper is to use a Color Descriptors which work on RGB and HSV color space after that this result is used by Improved K Mean with Steepest ascent (Gradient) algorithm. In which it used a heuristic local search algorithm that provide additional information about the solution. In this direction it gives the effective result of clustering that provide stability and performs better in global searching.

EXISTING SYSTEM:
In conventional clustering method centroid value is depending on Cluster value so its changed according to changing in cluster value so its a main problem. And it not provide a best centroid value, so to solve this problem an improved K means Algorithm is developed with heuristic local search algorithm. This Provides additional information. This gives the best centroid value and stability in clustering.

PROPOSED SYSTEM:
In K Mean clustering defined the value of k and initializes the K cluster center. K Means is a popular clustering algorithm which adopts an iterative refinement procedure to determine data partitions and to compute their associated centre of mass, called centroids , partition the objects to the nearest clustering centroid to form a cluster according to the Nearest-Neighbor rule, then compute the mean value of each cluster and makeit the new clustering centroid and in

K mean work in iterative manner and used a error rule Function E .the error function is as follows: E = i=1 k p ci (p-mi)2 (1) Where p denotes the object in space, mi is the centroid of cluster Ci. The smaller E is, the more similar within group data

SOFTWARE REQUIREMENTS:

Operating system Coding Language IDE

: : :

Windows XP Visual C# .Net Microsoft Visual Studio 2008

HARDWARE REQUIREMENTS: Processor Hard Disk RAM : : : Pentium IV 80 GB. 1GB.

You might also like