Hi there,

I have a problem with this task.

The Antirobot classifier must distinguish between robots and humans. Michael and Lisa wrote classifiers for a stream about which it is known that exactly half of the robots are there. Michael's classifier recognizes 80% of robots out of all robots, but finds (incorrectly) 40% of robots out of all humans. Lisa's classifier recognizes 60% of robots out of all robots, but counts 20% of humans as robots. If we assume that they wrote the same classifier that determines the "robotness" of the visit in the range from 0 to 1, but chose a different response threshold, what is the maximum f-score for this classifier?