BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//UNIFR/WEBMASTER//NONSGML v1.0//EN
CALSCALE:GREGORIAN
BEGIN:VEVENT
DTSTART;VALUE=DATE:20260313T131500
DTEND;VALUE=DATE:20260313T131500
UID:19446@agenda.unifr.ch
DESCRIPTION:This project presents a comparative study of three node ranking metrics – in-degree, PageRank, and Zermelo’s strength – in competitive directed networks. Through numerical simulations on different synthetic networks, the study evaluated their mutual correlations with respect to the predefined intrinsic strength of the nodes. Correlation analyses using Pearson and Spearman coefficients reveal that Zermelo’s strength provides the most accurate estimates across all network configurations, particularly in networks where participation is unequal. In-degree shows high sensitivity to network density and participation bias, while PageRank exhibits moderate performance with limited dependence on its teleportation probability parameter. This work is a basis for future analysis of real-world competitive data in professional tennis rankings.
SUMMARY:A Comparative Study of Node Ranking Metrics in Directed Networks
CATEGORIES:Autre
LOCATION:PER 08\, 0.51\, Chemin du Musée 3\, 1700 Fribourg
URL;VALUE=URI:https://agenda.unifr.ch/e/fr/19446
END:VEVENT
END:VCALENDAR