Content deleted Content added
GoingBatty (talk | contribs) clean up, typo(s) fixed: doesn’t → doesn't, appropiate → appropriate, ’s → 's, 583-617 → 583–617 |
grammar error |
||
Line 23:
<math>C(i,j)=\left ( \frac{\textstyle \sum_{h=1}^H M^h(i,j) \displaystyle}{\sum_{h=1}^H I^h(i,j)} \right )</math>
That is the entry <math>(i,j)</math> in the consensus matrix is the number of times points <math>i</math> and <math>j</math> were clustered together divided by the total number of times they were selected together. The matrix is symmetric and each element is defined within the range <math>[0,1]</math>. A consensus matrix is calculated for each <math>K</math> to be tested, and the stability of each matrix, that is how far the matrix is towards a matrix of perfect stability (just zeros and ones) is used to determine the optimal <math>K</math>. One way of quantifying the stability of the <math>K</math>th consensus matrix is examining
==Over-interpretation potential of the Monti consensus clustering algorithm==
| |||