MEASURE DIS SIMILARTY OF PERSON
Measuring Dis/Similarities Between Objects (Cells) In 'n'-Dimensional Space Cluster analysis works creating groups that have minimal variance (more similar) within and maximal dissimilarities between them. Cluster analysis can work either agglomeratively or divisively. Agglomerative methods begin with each individual cell representing a group (or cluster) and joins the two most similar to form a new cluster. This then repeats until all of the cells have been clustered into one group. I'll explain this more clearly with graphics below because it is the technique that I will be using classify my cells. Divisive methods start with the entire set of cells, and subdivides it to maximize the dissimilarity between the 2 newly formed groups. Distance Measures For Agglomerative Methods How do you tell how similar two cells are? It should be pretty obvious from the previous graphs that you can tell clusters apart from each other because the points in space are closer to e...
Hdh
ReplyDelete