Don't be fooled by the name — logistic regression is not for guessing numbers, but for answering yes or no questions!
Imagine it like this: Instead of drawing a straight line like linear regression does, picture drawing an "S"-shaped slide. The higher a thing lands on the slide, the more likely the answer is "yes." Lands low, and the answer is probably "no."
Examples:
So? Logistic regression turns clues into a probability (like 82% chance), then picks the most likely yes/no answer.