Fw: [分享] ML 作業

LFD Ex 1.12
大致上是問你: 如果你的朋友 (friend) 給你一些 training data, 並要你對 test data
作 prediction, "She says the target function is completely unknown", 在這樣的
setting 下你能給他怎麼樣的 guarantee?

(a) you can provide a hypothesis g that always yield good approximation to f.
(b) you can provide g that with high prob. yield good approximation to f.
(c) one of two things happen:
    1) you will produce g which, with high prob. yield good approx.
    2) you declare you cannot find a good g

甲生答曰:

According to the description, I would choose (a).
Because the pronoun used in the question is "she", providing a guarantee is
wisdom choice. There is no need for any uncertain words such as "probably",
"perhaps", "maybe". She will not be glad to hear them.

==

Bonus(2):
If we replace your "friend" in the exercise with your "foe", you need to be
extremely careful in the guarantee you can provide. Do you see any traps in the

setup and/or the choices?

Write down your own version of the guarantee that you'd provide to your foe,

and state the convincing reasons behind your own version.

乙生答曰:

I will escape from the foe to avoid being attacked by the foe.

==

.......... 整個笑到無法自己 XD

foe 好可怕.. 甲生好專業..

--