A Note on the Core of 2-Matching Games

Laura Sanità,Lucy Verberk

2024-11-28

Abstract:Cooperative 2-matching games are a generalization of cooperative matching games, where the value function is given by maximum-weight b-matchings, for a vertex capacity vector $b \leq 2$. We show how to separate over the core of 2-matching games in polynomial time, fixing a small flaw in the literature, and prove the existence of a compact extended formulation for it.

Computer Science and Game Theory,Discrete Mathematics

What problem does this paper attempt to address?

### Problems the paper attempts to solve

This paper mainly focuses on the core problems of cooperative 2 - matching games. Specifically, the paper solves the following two key problems:

1. **Core separation problem**: How to determine whether a given allocation belongs to the core of a 2 - matching game in polynomial time.

- **Background**: In cooperative games, the core is a set of allocation vectors that ensure no subset of players has the incentive to leave the grand coalition to form their own coalition. For 2 - matching games, the core separation problem refers to how to determine whether a given allocation satisfies the core conditions or find the subsets that violate these conditions.

- **Contribution**: The paper corrects a small error in the literature and proves that the core separation problem of 2 - matching games can be solved in polynomial time.

2. **Existence of a compact extended formula**: Whether there exists a compact extended formula to describe the core of 2 - matching games.

- **Background**: A compact extended formula is an effective method for describing polyhedra, which can significantly reduce the computational complexity. For some game problems, even if there is a polynomial - time separation algorithm, a compact extended formula may not exist.

- **Contribution**: The paper proves that there indeed exists a compact extended formula for the core of 2 - matching games.

### Overview of specific content

#### 1. Introduction
The paper first introduces the concept of cooperative 2 - matching games, which is a generalization of cooperative matching games. In 2 - matching games, the value function is given by maximum - weight b - matchings, where the maximum capacity \( b\leq2 \) for each vertex. The paper points out that Shapley and Shubik introduced the concept of cooperative matching games as early as 50 years ago.

#### 2. Core separation problem
The paper discusses in detail how to solve the core separation problem of 2 - matching games in polynomial time. The specific steps are as follows:
- **Cycle constraints**: By transforming the problem into the "tramp steamer problem", all cycle constraints can be checked in polynomial time.
- **Path constraints**: The paper points out an error in the path separation method proposed by Biró et al. and proposes a new method to correct this problem. The new method also depends on the "tramp steamer problem".

#### 3. Compact extended formula
To prove that there exists a compact extended formula for the core of 2 - matching games, the paper introduces a linear programming (LP) model. This model utilizes Seymour's theorem on the cycle cone of graphs and converts it into a compact LP form through the concept of flow.

### Conclusion
Through the above methods, the paper not only corrects the errors in the existing literature but also proves that the core separation problem of 2 - matching games can be solved in polynomial time and that there exists a compact extended formula for its core. This provides an important theoretical basis for the research of 2 - matching games.

### Related formulas
- **Value function**:
\[
\nu(S)=w(M_S)=\sum_{e\in M_S}w_e
\]
where \( M_S\subseteq E \) is the maximum - weight matching in the subgraph \( G[S] \).

- **Core conditions**:
\[
p(N)=\nu(N)
\]
\[
p(S)\geq\nu(S)\quad\text{for all }S\subset N
\]

- **Tramp steamer problem**:
\[
\max_{C\in C}\frac{w(C)}{p'(C)}
\]
where \( C \) is a cycle in the graph and \( p' \) is the edge weight transformed according to the allocation \( p \).

The correctness and readability of these formulas are guaranteed, ensuring that users can clearly understand the key content of the paper.