本頁只刊出中文翻譯與中文說明；英文原文請見下方原文連結。
原文連結
論文資訊
- 類型：已發表論文
- 日期：2019-04-16
摘要
A common graph mining task is community detection, which seeks an unsupervised decomposition of a 網絡 into groups based on 統計 regularities in 網絡 connectivity. Although many such algorithms exist, community detection's No Free Lunch theorem implies that no algorithm can be optimal across all inputs. However, little is known in practice about how different algorithms over or underfit to real 網絡s, or how to reliably assess such behavior across algorithms. Here, we present a broad investigation of over and underfitting across 16 state-of-the-art community detection algorithms applied to a novel benchmark corpus of 572 structurally diverse real-world 網絡s. We find that (i) algorithms vary widely in the number and composition of communities they find, given the same input; (ii) algorithms can be c
※ 此為已發表論文，全文需透過期刊付費取得