Rank Order Clustering

This method is depends on sorting the rows & columns of machine part incidence matrix. The rank order clustering was built up by King (1980). Steps of this algorithm is illustrated below:

Step 1

For each row of the machine part incidence matrix, assign binary weight & calculate the decimal equivalent

Step 2

Sort all rows of the binary matrix in decreasing order of the equivalent decimal weights.

Step 3

Repeat the preceding two steps for each of the column.

Step 4

Repeat the preceding steps till the location of each element in each row & column does not change.

A weight for each of rows i & column j are calculated as:

In the ultimate matrix generated by the ROC algorithm, clusters are recognized visually.