How statement and proof provenance work
The first chip identifies the source of the statement or construction; the second identifies the source of its local proof or verification.
- Literature-sourced: the exact statement appears in a cited source; only wording and notation differ.
- AI-adapted: a semantically identical restatement of literature-sourced material, modulo indexing, notation, and boundary cases adopted by the library.
- AI-generated: a genuinely novel statement formulated by AI, with no source for the claim itself.
These labels describe origin, not correctness: citations and verification chips remain separate evidence.
The greedy charge on each newly covered element is at most OPT divided by the remaining count
Statement
Let elements remain uncovered just before a weighted greedy set-cover step, and let be the cost of a fixed minimum cover. The chosen price per newly covered element is at most . Equivalently, if the elements are ordered by first coverage, ties within a round following the fixed order of the universe, then the -th element of this order receives charge at most , where is the size of the universe.
Facts & Assumptions
Given: A feasible weighted set-cover instance with universe of elements, listed sets with nonnegative rational costs, a run of the weighted greedy algorithm, a step of that run, and a minimum-cost cover with total cost .
While uncovered elements remain, the greedy algorithm chooses a listed set with positive newly covered count minimizing the ratio of its cost to that count, ties going to the smallest index in the input order; each set chosen in a round charges every element it newly covers the same amount, namely its cost divided by its newly covered count. All comparisons are exact in the rationals, and the listed family covers . (Weighted greedy set cover and element charges)
Every instance of the stated domain has at least one feasible solution, the set of values of feasible solutions is a finite nonempty set of nonnegative rationals, and its minimum in the minimization direction is attained; the optimum is therefore a well-defined nonnegative rational, denoted for this problem. (Optimization problems and approximation ratios)
Proof
Fix a step of the greedy run, let be the set of elements still uncovered just before it and the remaining count; let , , be the set of elements newly covered by the set chosen at this step, so the chosen price per newly covered element is for that set's cost . Put for every set of the fixed minimum-cost cover of [F2]. Since covers and hence , every element of lies in at least one ; summing the counts therefore counts each element of at least once, so .
Let . Then is nonempty because , and , while because all costs are nonnegative and . The -weighted average of the ratios over equals and is therefore at least the minimum of those ratios, so that minimum is at most .
Each is a listed set of the instance and has positive newly covered count at this step, so it is a candidate in the greedy choice; by [F1] the chosen set minimizes the ratio among all candidates, so the chosen price per newly covered element satisfies . Every element newly covered at this step receives the charge , hence a charge of at most .
Order the elements of by the round in which they are first covered, breaking ties within a round by the fixed order of , and let be the -th element of this order, covered in a round that begins with uncovered elements. Before that round exactly elements are already covered, all of them earlier than in the order, so , that is, ; since the charge of is at most by step 3.1 and , it is at most .
Consequently the price chosen in any greedy step with uncovered elements is at most , and the -th element in first-coverage order receives charge at most for every ; both bounds are ordinary inequalities of nonnegative rationals with positive denominators, so no division by zero occurs.
Depends on
Used by
Dependency tree · two levels
4 results within two dependency steps of this one, each drawn at its shortest distance from it. An arrow runs from a result to what uses it, so the chart reads left to right and ends at this result, which carries a heavier outline. Every node is a link to that result. Click elsewhere on the chart to enlarge it.
Sources
- Williamson and Shmoys, The Design of Approximation Algorithms, §1.6 Fact 1.10 and its proof, printed p. 25 (standard reference, not scraped)
- Ghaffari, Advanced Algorithms, Lecture 1: Approximation Algorithms I, §2.1 Theorem 3, PDF pp. 2–3 (standard reference, not scraped)