Assume I have the below dataset, features M and N are numerical, label is binary.
| M | N | Label |
|---|---|---|
| 2 | 3 | y |
| 6 | 1 | y |
| 1 | 12 | y |
| 3 | 9 | y |
| 11 | 15 | n |
| 7 | 13 | n |
| 4 | 8 | n |
| 9 | 10 | n |
My decision tree with binary split only has 1 root node and 2 leaf nodes (no internal node)
I can use M or N to make a split.
What is the size(cardinality) of Hypothesis Space |H| ? Could you explain detail ?
