An MSSP-Like Problem




Hello, Everyone,

We need advice on an optimization problem
that we are trying to solve. The problem is stated
as follows.

Given a set of m bins with capacities c_1, c_2,
...., c_m, and a set of n items with positive integral
weights w_1, w_2, ..., w_n, the objective is to assign
the items to the bins so that the sum of 'effective
weights' is maximized. The 'effective weight' with
respect to a bin is defined as the minimum of 'the
total weights of the items assigned to the bin' and
'the capacity of the bin'. Each item can either stay
unassigned or be assigned to exact one bin. That is,
the goal is to

maximize Sum_i(Min(Sum_j(w_j * x_ij), c_i))

subject to (1) Sum_i(x_ij) <= 1 for j = 1, 2, ..., n,
and (2) x_ij is either 0 or 1 for i = 1, 2, ..., m,
and j = 1, 2, ..., n.

This problem is similar to the Multiple Subset
Sum Problem (MSSP). However, the problem stated here
allows the sum of the weights of the items assigned to
a bin to exceed the capacity of the bin, and then it
uses the Min() function to balance the gain of
such assignment.

What we want to know is the complexity of this
problem. Please kindly share your opinions with us.
Thanks in advance.

Regards,
Charles


.



Relevant Pages

  • An MSSP-Like Problem
    ... weights' is maximized. ... respect to a bin is defined as the minimum of 'the ... 'the capacity of the bin'. ... Sum Problem. ...
    (comp.theory)
  • Re: Injuries in 19 yr old women
    ... I would bin this one. ... Heavy weights aren't necessarily a problem. ... The one instance of a college rower that I know, ... Dead Lift, no Cleans, no Bent Over Rowing. ...
    (rec.sport.rowing)
  • Re: Histogram with weighting factors?
    ... If you can quantize your original data so that there's one data point ... per bin then you can just multiply your weights by your histogram. ...
    (comp.soft-sys.matlab)
  • Re: weighted variance comparisons
    ... you gave and checked it with using weights of 1 so I am happy with my ... peterp does not seem to me to take any account of the variance of the ... data within each weighting bin. ... So, to calculate the total variance, you not only need the ...
    (sci.stat.math)
  • Whut IZ dey talkin abowt?
    ... Shoooore, she bin on da 'puter, but not doin' OWER typin. ... She's lukkin at sum weeeerd stuff. ... Apparently she an de guy-humin sed sumpin' abowt offerin' sum uther ... humin sumpin, an' de uther humin acseptid de offer. ...
    (rec.pets.cats.community)