User Tools

Site Tools


en:non-hier

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
Last revision Both sides next revision
en:non-hier [2019/03/22 21:58]
David Zelený
en:non-hier [2019/04/06 18:48]
David Zelený [K-means (non-hierarchical classification)]
Line 7: Line 7:
 [[{|width: 7em; background-color:​ white; color: navy}non-hier_exercise|Exercise {{::​lock-icon.png?​nolink|}}]] [[{|width: 7em; background-color:​ white; color: navy}non-hier_exercise|Exercise {{::​lock-icon.png?​nolink|}}]]
  
-==== kmeans ==== +This is a non-hierarchical ​agglomerative clustering algorithmbased on Euclidean distances among samples and using an iterative algorithm to find the solution. It minimizes the total error sum of squares (TESS), the same objective function as in the case of Ward’s algorithm. The number of clusters (k) is defined by the userOther than Euclidean distance can be usedbut they need to be converted into metric distances ​and submitted to PCoAFor example, in the case of Bray-Curtis distance, which is not metric, one may calculate square-rooted Bray-Curtis distances (which are metric), submit them to PCoA, and then use all PCoA axes as the input matrix in K-means method instead ​of the raw data. The K-means algorithm, similarly to other iterative methods (like NMDS) can get trapped in local minima, and it may be useful to repeat the analysis many times and choose the solution with the lowest overall TESS.
-Non-hierarchical ​classification, using method ​of //k// meansNon-hierarchical methods are overlookedeven if they give ecological interesting ​and relevant resultsYou need to a priori set up the number ​of clusters you want the data divide into.+
  
-<code rsplus> 
-cluster.kmeans <- kmeans (dis, centers = 5) 
-cluster.kmeans$cluster 
-</​code>​ 
-<​file>​ 
- ​1 ​ 2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26  
- ​4 ​ 4  4  4  5  5  1  2  4  4  5  4  4  4  5  2  1  2  4  1  4  5  2  2  4  3  
-27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52  
- ​2 ​ 2  2  5  5  1  1  1  1  1  2  4  4  4  2  2  2  5  4  4  4  1  5  1  1  2  
-53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78  
- ​2 ​ 3  5  3  3  3  3  3  3  3  3  5  3  3  3  3  3  2  2  2  2  2  3  4  4  5  
-79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97  
- ​3 ​ 3  3  5  3  3  3  5  4  4  4  5  2  1  2  5  2  1  1 
-</​file>​ 
en/non-hier.txt · Last modified: 2019/04/06 18:53 by David Zelený