Efficient Pairwise Classification and Ranking
| Type of publication: | Techreport |
| Citation: | jf:TUD-KE-2007-03 |
| Number: | TUD-KE-2007-03 |
| Year: | 2007 |
| Institution: | TU Darmstadt, Knowledge Engineering Group |
| URL: | /publications/reports/tud-ke-2007-03.pdf |
| Abstract: | Pairwise classi cation is a class binarization procedure that converts a multi-class problem into a series of two-class problems, one problem for each pair of classes. While it can be shown that for training, this procedure is more efficient than the more commonly used one-against-all approach, it still has to evaluate a quadratic number of classifi ers when computing the predicted class for a given example. In this paper, we propose a method that allows a faster computation of the predicted class when weighted or unweighted voting are used for combining the predictions of the individual classi ers. While its worst-case complexity is still quadratic in the number of classes, we show that even in the case of completely random base classi ers, our method still outperforms the conventional pairwise classifi er. For the more practical case of well-trained base classi ers, its asymptotic computational complexity seems to be almost linear. We also propose a method for approximating the full class ranking, based on the Swiss System, a common scheme for conducting multi-round chess tournaments. Our results indicate that this adaptive scheme o ffers a better trade-off between approximation quality and number of performed comparisons than alternative, fi xed schemes for ordering the evaluation of the pairwise classi fiers. |
| Keywords: | |
| Authors | |
|
Attachments
|
|
|
|
|
Topics
|
|
|
|
