Hi,

Hope everyone is doing great!

I am thinking about modify the algorithm to secretary problem and
apply it to get airline purchase.

Also, I know one area where price prediction has been more extensive
is auctions. Sellers can benefit from algorithms that predict the
final selling price of an item based on similar past auctions. Ghani
[2005] put together three types of algorithms to approach this
problem: regression, decision trees, and neural networks.

Right now, I think secretary problem is more akin to airline ticket
purchase problem. So I am going to work on the algorithm to secretary
problem for my airline ticket purchase.

Any body has any suggestions or ideas or advice on my ideas, please?
Specifically, I want to know listen to your experts' opinion on the
advantage/ disadvantage of the algorithm to the secretary problem, the
regression, decision trees, and neural networks algorithm for auction.
I right now just have a very general idea of the advantage /
disadvantage. It will be wonderful if I can get some advice here.

Finally, I cannot think of a theoretical way in explaining about why
secretary problem is more akin to airline ticket purchase problem. Any
one has any idea, please? I just think it is intuitively.

Thanks.



Best regards,
Sarah

They are not quite the same. If you reject a secretary candidate,
they are out of the game immediately (the sequential part of the
problem). I can have several airline tickets available to me at the
same time, but they are in the candidate set only for a limited
time.