You can edit almost every page by Creating an account. Otherwise, see the FAQ.

Mountain Clustering Method

From EverybodyWiki Bios & Wiki



Mountain Clustering Method[edit]

Mountain Clustering Method is used to obtain cluster centers during image segmentation to classify the pixels of an image. The Mountain Clustering Method finds the clustering center by computer the density of the mountain function, the highest value of the mountain function will be identified as the clustering center.

History[1][edit]

The term "Mountain Clustering" was first introduced by R.R.Yager in 1944.

Algorithms[edit]

To determine cluster center using Mountain Clustering Method, there are three steps:

The first step is to forming data space grid. The data space grid is a 2-d data where the intersections of the grid lines are represents for cluster centers, denoted as a set V.

The second step is to find the density measure, which is the height of the mountain function at given point V:

To find density measure m.

where xi is the ith data point and s is a constant. In the equation, xi represent the height of the mountain function at v. The mountain function is a measure of data density because the more data around, the value of the m tends to be greater. Since m is tends to be higher value if more data points are located nearby, and lower if fewer data points are around. The constant m also represents smoothness of the mountain function.

The third step is to find the greatest value of the cluster center. To find the greatest value of the cluster enter, first find a random cluster center c1. To find the next cluster center requires to eliminating the c1. This requires to change the mountain function by subtracting a scaled Gaussian function at c1:

To find the greatest value of the cluster center.

After finding the c2, this process of finding the next cluster centers continues until a sufficient number of cluster centers is attained.


This article "Mountain Clustering Method" is from Wikipedia. The list of its authors can be seen in its historical and/or the page Edithistory:Mountain Clustering Method. Articles copied from Draft Namespace on Wikipedia could be seen on the Draft Namespace of Wikipedia and not main one.

  1. "Approximate clustering via the mountain method - IEEE Journals & Magazine". ieeexplore.ieee.org. Retrieved 2018-04-23.