We present a \emph{deterministic exact algorithm} for the \emph{minimum $k$-cut problem} on simple graphs. Our approach combines the \emph{principal sequence of partitions (PSP)}, derived canonically from ideal loads, with a single level of \emph{Kawarabayashi--Thorup (KT)} contractions at the critical PSP threshold~$λ_j$. Let $j$ be the smallest index with $κ(P_j)\ge k$ and $R := k - κ(P_{j-1})$. We prove a structural decomposition theorem showing that an optimal $k$-cut can be expressed as the level-$(j\!-\!1)$ boundary $A_{\le j-1}$ together with exactly $(R-r)$ \emph{non-trivial} internal cuts of value at most~$λ_j$ and $r$ \emph{singleton isolations} (``islands'') inside the parts of~$P_{j-1}$. At this level, KT contractions yield kernels of total size $\widetilde{O}(n / λ_j)$, and from them we build a \emph{canonical border family}~$\mathcal{B}$ of the same order that deterministically covers all optimal refinement choices. Branching only over~$\mathcal{B}$ (and also including an explicit ``island'' branch) gives total running time $$ T(n,m,k) = \widetilde{O}\left(\mathrm{poly}(m)+\Bigl(\tfrac{n}{λ_j}+n^{ω/3}\Bigr)^{R}\right), $$ where $ω< 2.373$ is the matrix multiplication exponent. In particular, if $λ_j \ge n^{\varepsilon}$ for some constant $\varepsilon > 0$, we obtain a \emph{deterministic sub-$n^k$-time algorithm}, running in $n^{(1-\varepsilon)(k-1)+o(k)}$ time. Finally, combining our PSP$\times$KT framework with a small-$λ$ exact subroutine via a simple meta-reduction yields a deterministic $n^{c k+O(1)}$ algorithm for $c = \max\{ t/(t+1), ω/3 \} < 1$, aligning with the exponent in the randomized bound of He--Li (STOC~2022) under the assumed subroutine.
翻译:本文针对简单图上的最小k割问题提出了一种确定性精确算法。我们的方法结合了从理想负载规范导出的主划分序列,以及在关键PSP阈值λ_j处进行的单层Kawarabayashi-Thorup收缩。令j为满足κ(P_j)≥k的最小索引,且R := k - κ(P_{j-1})。我们证明了一个结构分解定理:最优k割可表示为层级(j-1)边界A_{≤j-1},加上恰好(R-r)个值至多为λ_j的非平凡内部割,以及P_{j-1}各部分内部的r个单点隔离("孤岛")。在此层级,KT收缩产生总规模为Õ(n/λ_j)的核心子图,并由此构建同阶数的规范边界族B,该族能确定性覆盖所有最优细化选择。仅对B进行分支(同时包含显式"孤岛"分支)得到的总运行时间为:
T(n,m,k) = Õ(poly(m) + (n/λ_j + n^{ω/3})^R),
其中ω < 2.373为矩阵乘法指数。特别地,若对某个常数ε > 0满足λ_j ≥ n^ε,则可获得确定性亚n^k时间算法,其运行时间为n^{(1-ε)(k-1)+o(k)}。最后,通过简单元归约将我们的PSP×KT框架与小λ精确子程序相结合,得到确定性n^{ck+O(1)}算法,其中c = max{t/(t+1), ω/3} < 1,该指数与He--Li(STOC 2022)在假设子程序下的随机化界限中的指数保持一致。