5.17.3.B. Unsupervised classification

Try these first

1. 1.

   What is the main focus of unsupervised classification?

   Hint

   Think about what kind of data this classification relies on.

2. 2.

   Name one algorithm used in unsupervised classification.

   Hint

   Consider popular clustering methods.

3. 3.

   What does unsupervised classification rely on?

   • Training samples
   • Spectral properties
   • Ground truth data

   Hint

   Consider the absence of training in this method.

4. 4.

   True or False: ISODATA is a clustering algorithm that requires training samples.

   • True
   • False

   Hint

   What type of classification is ISODATA associated with?

5. 5.

   Suppose you have satellite imagery of a forest area with varying tree types. Describe how you would apply unsupervised classification and what considerations you would account for.

   Hint

   Think about how diversity in the forest might impact classification.

6. 6.

   What strategies could you use if your unsupervised classification resulted in ambiguous clusters?

   Hint

   What might you look for to clarify your clusters?