Is there a strategy for combining the German tank and secretary problem?

  • Context: Graduate 
  • Thread starter Thread starter Gerenuk
  • Start date Start date
  • Tags Tags
    German Tank
Click For Summary
SUMMARY

The discussion focuses on the strategy for combining the German tank problem and the secretary problem, specifically in scenarios where the total number of tanks is unknown and the goal is to maximize the serial number of the selected tank. The participants explore the implications of a uniform distribution of serial numbers and the potential for a normal distribution in related cases. A survey article from 1983 is suggested as a resource for further insights into this combined problem.

PREREQUISITES
  • Understanding of the German tank problem and its implications.
  • Familiarity with the secretary problem and optimal stopping theory.
  • Knowledge of statistical distributions, particularly uniform and normal distributions.
  • Ability to analyze problems involving unknown sample sizes and distributions.
NEXT STEPS
  • Read the 1983 survey article referenced in the Wikipedia entries for deeper insights.
  • Research optimal stopping strategies in the context of the secretary problem.
  • Explore statistical methods for estimating unknown distributions, focusing on uniform and normal cases.
  • Investigate applications of the German tank problem in real-world scenarios to understand its practical implications.
USEFUL FOR

Mathematicians, statisticians, operations researchers, and anyone interested in optimal decision-making strategies under uncertainty.

Gerenuk
Messages
1,027
Reaction score
5
Is it possible to give a strategy for the combined problem of
http://en.wikipedia.org/wiki/German_tank_problem
and
http://en.wikipedia.org/wiki/Secretary_problem

So I'm observing tanks with serial numbers and I'm allowed only the keep the last one. I also don't know how many there are in total. My task is to (increase change of getting the highest number) or (achieve the highest average serial number on the tank I'm keeping). The distribution of serial numbers is uniform. Another case would be if it's normal.

Basically it is the secretary problem with an unknown number of applicants, but moreover this unknown number if applicants has to be estimated given the observed incoming samples. The assumed distribution of all samples could be uniform or maybe also normal.

Has anyone heard of an article where this problem is considered? Or how do I find such an article?
 
Physics news on Phys.org
Gerenuk said:
Basically it is the secretary problem with an unknown number of applicants

Interesting problem. Have you tried the survey article that was listed in the wikipedia article references? It's from 1983 but might say if the same problem is solved under a different name.

Also, how could serial numbers be drawn from the normal distribution?
 
bpet said:
Also, how could serial numbers be drawn from the normal distribution?
Actually I meant this as two different cases. Serial numbers from a uniform distribution. And on the other hand some other cardinal value from a normal distribution like the height of people (with apriori unknown mean height).

I just downloaded the survey article. Hope I find some time to go through it soon.
 

Similar threads

  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 13 ·
Replies
13
Views
3K
  • · Replies 24 ·
Replies
24
Views
3K
  • · Replies 8 ·
Replies
8
Views
2K
  • · Replies 24 ·
Replies
24
Views
7K
  • · Replies 3 ·
Replies
3
Views
3K
  • · Replies 11 ·
Replies
11
Views
2K
  • · Replies 8 ·
Replies
8
Views
5K