Estimating Mixed Memberships with Sharp Eigenvector Deviations (JASA, 2021) People belong to multiple overlapping communities. Can we detect these communities from the structure of a social network? We show how one can consistently infer these communities, and the affinities that each person feels towards each community. These follow from a new sharp bound on eigenvectors, which may be of general
リリース、障害情報などのサービスのお知らせ
最新の人気エントリーの配信
処理を実行中です
j次のブックマーク
k前のブックマーク
lあとで読む
eコメント一覧を開く
oページを開く