Show that the MINIMUM CUT PROBLEM is the dual program of the MAXIMUM FLOW PROBLEM



Show that the MINIMUM CUT PROBLEM is the dual program of the MAXIMUM FLOW PROBLEM

Let $(G=(V,E), s,t,c)$ be a network.
Let consider problem (4.1) maximize $\sum_{v: (s,v) \in E} f(s,v)$ with constraints $\sum_{u:(u,v) \in E} f(u,v) = \sum_{w : (v,w) \in E} f(v,w)$ for all $v \in V-\{s,t\}$, $f(u,v) \leq c(u,v)$ for all $(u,v) \in E$ and $f(u,v) \geq 0 $ for all $(u,v) \in E$.
Let $P$ be a set of all paths from $s$ to $t$ and construct problem (4.2) maximize $\sum_{p\in P} x_p$ with $\sum_{p\in P:(u,v) \in P}x_p \leq c(u,v)$ for all $(u,v) \in E$ and for all $p\in P$, and $x_p \geq 0 $ for all $p \in P$.
I will neither prove nor present Flow Decomposition theorem, but I believe it is true, by this theorem, the assumption (4.1) has feasible solution implies (4.2) has feasible solution and they have the same cost.
Suppose that $x_p$ is feasible solution for (4.2), then $f(u,v):=\sum_{p \in P:(u,v) \in P}x_p$ is a feasible solution of (4.1) and $\sum_{v:(u,v)\in E} f(u,v) = \sum_{p\in P: v \in p} x_p = \sum_{w : (v,w) \in E} f(v,w)$.
Therefore (4.1) and (4.2) are equivalent.
(4.2) has dual problem (4.3) minimize $\sum_{(u,v) \in E} c(u,v) y_{u,v} $ with constraitns $\sum_{(u,v) \in P} y_{u,v} \geq 1$ for all $p \in P, y_{u,v} \geq 0 $ for all $(u,v) \in E$.
$y_v$ can be defined as $\sum_{(v,w) \in E} y_{v,w}$.
Then (4.3) has form minimize $\sum_{(u,v) \in E} c(u,v) y_{u,v}$ subject to $y_v + y+s_v \geq 1 $ for all $v $ such that $(s,v) \in F$, $y_v-y_u+y_{u,v} \geq 0$ for all $(u,v) \in E$ with $u\neq s, v\neq t$, $-y_u+y_{u,t} \geq 0 $ for all $u$ with $(u,t) \in E$.

Leave a Comment