SPSS-System Clustering Hand Calculation Practice

I. Systematic clustering characteristics

It is possible to point out the classification results from coarse to fine series Example: sample point set:  ={w1,w2,…,w6,w7}

"Pedigree diagram": fi for "aggregation index" — indicating the level of aggregation

Second, the similarity measure between sample points-clustering

Definition: Distance d( , ):     R+ ① d(x,y)≥0 x,y  (positive definiteness) d(x,y) = 0 if and only if x=y ② d(x,y) = d(y,x) (symmetry) ③ d(x,y)  d(x,z) + d(z,y) (triangular inequality)

Three, the similarity measure between classes - aggregation index

Question: If there are two types of G1, G2, how to measure the distance between them? —"Aggregation Index" phylogenetic clustering is a stepwise process:

1. Shortest distance method:

2. Longest distance method:

3. Center of Gravity:

4. Class average method:

Four. Practice