Finding the Bad Ones: Identifying Unsuccessful Warship Designs

• Activity Number: 608
• Type: Topic Contributed
• Date/Time: Thursday, August 4, 2011 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics in Defense and National Security
• Author(s): Steffen Liebscher*+ and Thomas Kirschstein and Claudia Becker
• Companies: Martin-Luther-Universität Halle-Wittenberg and Martin-Luther-Universität Halle-Wittenberg and Martin-Luther-Universität Halle-Wittenberg
• Keywords: warship scoring ; outlier identification ; cluster detection

Abstract

Inspired by the book "The World's Worst Warships" by Preston (2002) we try to identify unsuccessful warship designs by means of advanced outlier detection procedures. Therefore, we generate a data set covering information on general characteristics, offensive/defensive power and sensor capabilities of more than 800 different classes of (surface) warships from the cold war era to the current era (including planned future classes). On this high-dimensional data set we apply two approaches, one based on self-organizing maps (i.e. a special kind of artificial neural network, Liebscher et al. 2011), the other one based on minimum spanning trees (Kirschstein et al., 2011). Both methods are modified to make them suitable for outlier identification (and to reveal cluster structures). The results show that our methods are indeed able to identify "unsuccessful" designs.