Estimating Attributed Central Orders
Lists of ordered objects are widely used as representational forms. Such ordered objects include Web search results or best seller lists. In spite of their importance, the methods of processing orders have received little attention. However, research concerning object ordering is becoming more common. Some researchers have developed various methods to perform almost the same task: a learning function used for sorting objects from examples of ordered sequences. We call this task the estimation of Attributed Central Orders (ACO for short). The performance of this task is useful for sensory surveys, information retrieval, or decision making. We performed a survey of such methods, empirically compared the methods’ properties, and discuss their merits and demerits.
KeywordsEstimation Task Learning Function Attribute Noise Representational Form Sample Order
- 3.Kazawa, H., Hirao, T., Maeda, E.: Order SVM: A kernel method for order learning based on generalized order statistic. The IEICE Trans. on Information and Systems, pt. 2 J86-D-II, 926–933 (2003) (in Japanese) [An English version will appear in “Systems and Computers in Japan” Wiley Periodicals, Inc.]Google Scholar
- 4.Herbrich, R., Graepel, T., Bollmann-Sdorra, P., Obermayer, K.: Learning preference relations for information retrieval. In: ICML 1998 Workshop: Text Categorization and Machine Learning, pp. 80–84 (1998)Google Scholar
- 5.Kamishima, T., Akaho, S.: Learning from order examples. In: Proc. of The IEEE Int’l Conf. on Data Mining, pp. 645–648 (2002)Google Scholar
- 6.Akaho, S., Kamishima, T.: A statistical approach for learning from order examples linear models. In: Proc. of the 16th Annual Conference of JSAI (2002) (in Japanese)Google Scholar