We investigate using gradient descent methods for learning ranking functions; we propose a simple probabilistic cost function, and we introduce RankNet, an implementation of these...
Christopher J. C. Burges, Tal Shaked, Erin Renshaw...
Classification, which involves finding rules that partition a given da.ta set into disjoint groups, is one class of data mining problems. Approaches proposed so far for mining cla...
Abstract Image processing, specifically Automatic Target Recognition (ATR) in Synthetic Aperture Radar (SAR) imagery, is an application area that can require tremendous processing ...
Ruth Sivilotti, Young Cho, Wen-King Su, Danny Cohe...
Various data mining applications involve data objects of multiple types that are related to each other, which can be naturally formulated as a k-partite graph. However, the resear...
Bo Long, Xiaoyun Wu, Zhongfei (Mark) Zhang, Philip...
This paper proposes an incremental multiple-object recognition and localization (IMORL) method. The objective of IMORL is to adaptively learn multiple interesting objects in an ima...