Chapter 2: Threshold values

jumanamp jumanamp

I am stuck with understanding how are the thresholds being calculated in stump.py. In figure 2.py, it says manually written thresholds as 1.65 and 1.75. Why are we doing that? Can I get some insight here?

As I print, all possible threshold values are the copies of all data points for all features. But, I just want to know why we do this so.

You
Preview: