We introduce a new class of objectives for optimal transport computations of datasets in high-dimensional Euclidean spaces. The new objectives are parametrized by $\rho \geq 1$, and provide a metric space $\mathcal{R}_{\rho}(\cdot, \cdot)$ for discrete probability distributions in $\mathbb{R}^d$. As $\rho$ approaches $1$, the metric approaches the Earth Mover's distance, but for $\rho$ larger than (but close to) $1$, admits significantly faster algorithms. Namely, for distributions $\mu$ and $\nu$ supported on $n$ and $m$ vectors in $\mathbb{R}^d$ of norm at most $r$ and any $\epsilon > 0$, we give an algorithm which outputs an additive $\epsilon r$-approximation to $\mathcal{R}_{\rho}(\mu, \nu)$ in time $(n+m) \cdot \mathrm{poly}((nm)^{(\rho-1)/\rho} \cdot 2^{\rho / (\rho-1)} / \epsilon)$.
翻译:我们针对高维欧几里得空间中数据集的最优传输计算,引入了一类新的目标函数。该目标函数由参数 $\rho \geq 1$ 决定,并为 $\mathbb{R}^d$ 中的离散概率分布提供了一个度量空间 $\mathcal{R}_{\rho}(\cdot, \cdot)$。当 $\rho$ 趋近于 $1$ 时,该度量趋近于推土机距离,但对于大于(但接近)$1$ 的 $\rho$ 值,其算法速度显著加快。具体而言,对于支撑在 $\mathbb{R}^d$ 中至多 $n$ 和 $m$ 个向量(范数不超过 $r$)上的分布 $\mu$ 和 $\nu$,以及任意 $\epsilon > 0$,我们给出了一种算法,能在时间 $(n+m) \cdot \mathrm{poly}((nm)^{(\rho-1)/\rho} \cdot 2^{\rho / (\rho-1)} / \epsilon)$ 内输出 $\mathcal{R}_{\rho}(\mu, \nu)$ 的加性 $\epsilon r$-近似值。