Decision Trees - Bayes Rule Example

แชร์
ฝัง
  • เผยแพร่เมื่อ 19 ธ.ค. 2024

ความคิดเห็น • 2

  • @Whatdoyouwant1211
    @Whatdoyouwant1211 9 หลายเดือนก่อน

    sorry, i hope u reply. Yes we should pick a smaller expected value. So -1600 is smaller than -1575 right ???

    • @ISE_Post
      @ISE_Post  9 หลายเดือนก่อน

      We always pick the largest value (-1,575 > -1,600). In cases where the values represent cost, which is the case here, it implies less inspection cost over time. If the values represent profit, we will see positive values and larger values mean more profit. So always pick the largest value when you want to maximize the expected value of a decision tree. Also, remember, this is not for a one time decision. It is for setting policies because the decision will be applied to many many inspection cases and we want the average cost per inspection to be as low as possible.