Wasserstein distance in Optimal transport

最新推荐文章于 2025-02-26 20:57:11 发布

Aiqz

最新推荐文章于 2025-02-26 20:57:11 发布

阅读量1.3k

点赞数 1

文章标签：算法

本文链接：https://blog.csdn.net/qq_36903891/article/details/106557647

版权

Wasserstein Distance

Optimal transport

1. Notations

Consider two probability measures µ and ν defined on measure spaces X and Y . In most applications X and Y are subsets of $\mathbb{R}^d$ and µ and ν have density functions which we denote by $I_0$ and $I_1$ , $d\mu(x)=I_0(x)dx$ and $dv(x) = I_1(x)dx$ , (originally representing the height of a pile of soil/sand and the depth of an excavation).

2. Monge’s formulation

Monge’s optimal transportation problem is to find a measurable map f : X → Y that pushes µ onto ν and minimizes the following objective function,
$M(\mu,v)=inf_{f\in MP}\int_xc(x,f(x))d\mu(x)$
Where $c:X\times Y\rightarrow \mathbb{R}^+$ , is the cost functional, and $MP=\{f:X\rightarrow Y|f_\#\mu=v\}$ represents the pushforward of measure µ and is characterized as, $\int_{f^{-1}(A)} d \mu(x)=\int_{A} d \nu(y)$ for any measurable $A\subset Y$ .

在这里插入图片描述

Simply put, the Monge formulation of the problem seeks the best pushforward map that rearranges measure µ into measure ν while minimizing a specific cost function.

Drawback:

Nonlinear with respect to f(x)
For certain measures the Monge’s formulation of the optimal transport problem is illposed; in the sense that there is no transport map to rearrange one measure to another. For instance, consider the case where µ is a Dirac mass while ν is not.

3. Kantorovich’s formulation

Kantorovich’s formulation alleviates this problem by finding the optimal transport plan as opposed to the transport map. Kantorovich formulated the transportation problem by optimizing over transportation plans, where a transport plan is a probability measure $\gamma\in P(X \times Y)$ with marginals µ and ν. The quantity $\gamma(A,B)$ tells us how much ‘mass’ in set A is being moved to set B. Let $\Gamma(\mu,v)$ be the set of all such plans. Kantorovich’s formulation can then be written as,
$K(\mu, \nu)=\min _{\gamma \in \Gamma(\mu, \nu)} \int_{X \times Y} c(x, y) d \gamma(x, y)$
Note that unlike the Monge problem, in Kantorovich’s formulation the objective function and the constraints are linear with respect to $\gamma (x,y)$ . Moreover, Kantorovich’s formulation is in the form of a convex optimization problem.

在这里插入图片描述

The Kantorovich problem is especially interesting in a discrete setting, that is for probability measures of the form $\mu=\sum_{i=1}^{M} p_{i} \delta_{x_{i}}$