We Have Constraints

• We have a relatively small number of examples (e.g. “a”-pictures).

Therefore, the classifier must “learn” a “typical” from the given examples.

• We therefore want to “train” the classifier by showing it the “a” and non-”a” pictures and expect it to generalize.

(Thus, we cannot expect the final classifier to be always correct!)

השקופית הקודמת

השקופית הבאה

חזור אל השקופית הראשונה

הצג גירסה גרפית