YSBAT     Unsupervised Classification     03-13-07

- Explain the basic logic behind histogram peak analysis, including the difference between broad and fine generalization

- Under histogram peak analysis, explain the affect of limiting the number of clusters in the analysis

- Describe the two iterative self-organizing techniques ISODATA and ISOCLUST, highlighting their different "seeding" procedures, the decision rule they each apply, and the steps that are repeated in their “iterations”

- Compare what makes a cluster "significant" in the histogram peak analysis vs the iterative self organizing techniques.

- Explain 2 main differences between supervised and unsupervised classification