@techreport{oai:ipsj.ixsq.nii.ac.jp:00189227, author = {大久保, 壮浩 and 土中, 哲秀 and 小野, 廣隆 and Masahiro, Okubo and Tesshu, Hanaka and Hirotaka, Ono}, issue = {6}, month = {May}, note = {与えられた単純無向グラフ G = (V,E) の頂点集合 V の分割 {C₁,C₂,...,Ck} をグラフ分割という.グラフ G における頂点間の距離に基づき分割の社会的効用を定義し,これを評価尺度としたグラフ分割をグラフの最適社会効用分割と呼ぶ.一般に,グラフが与えられたときその最適社会効用分割を求めることは NP 困難である.本論文では,対象とするグラフを木に限定した場合,O (△²n) 時間で最適社会効用分割が得られることを示す.ただし,n は頂点数,△ は最大次数である., Given a graph G = (V,E), a partition {Ci, G2,...,Ck} of V is called a graph partition. We consider a social welfare of a graph partition of G based on distances between vertices in G, and define a graph partition with optimal social welfare. Finding a graph partition with optimal social welfare is known to be NP-hard in general. In this paper, we show that a graph partition with optimal social welfare of a given tree can be computed in O (△²n) time, where n is the number of vertices and △ is the maximum degree.}, title = {社会的距離に基づく木の最適分割}, year = {2018} }