E.B. Kong and T.G. Diettrich. Probability estimation via error-correcting output coding. Banff,Canada, 1997. Int. Conf. on Artificial Intelligence and soft computing. http://www.cs.orst.edu/ tgd/cv/pubs.html.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....has for L1 ECOC. The original Hamming Distance decoding strategy (HD ECOC) was shown in [20] 13 to be analogous to majority vote over the classes. Besides L1 ECOC, LS ECOC has been the most extensively investigated. The justification for LS ECOC in terms of probability estimation was reported in [21], and in [15] LS ECOC was extended by incorporating ridged regression when is small. In [15] it was also shown that for certain geometrical arrangements of patterns Cen ECOC is to be preferred to LS ECOC. From the perspective of computational learning theory, it was shown in [2] that a ....

E.B. Kong and T.G. Diettrich. Probability estimation via error-correcting output coding. In Int. Conf. of Artificial Inteligence and soft computing, Banff,Canada, 1997.

....where the outputs of the di erent predictors are non linearly combined through a gating network; ensemble constructed by subsets of input features [10] where each predictor selects a group of the input features. This paper focuses on Error Correcting Output Coding (ECOC) decomposition methods [16, 17, 32, 31, 28] and in particular on the factors a ecting the e ectiveness of these ensemble methods. Error correcting output codes [8] originally proposed to enhance the reliability of the transmission of binary signals through a noisy channel [35] have been successfully used in the framework of ....

....codes [52] In fact, using codewords for coding the classes suggest the introduction of codes with error recovering abilities. Kong and Dietterich showed that ECOC techniques can also provide class probability informations, through the solution of an over constrained system of linear equations [31]. An interesting extension of this approach presented by Schapire, consists of the combination of error correcting output codes with boosting techniques [19] this ensemble method shows good performances on di erent benchmark machine learning problems [49] From a statistical standpoint ECOC ....

B.E. Kong and T.G. Dietterich. Probability estimation via error correcting output coding. In IASTED International Conference: Arti cial Intelligence and Soft Computing, Ban, Canada.

....where the outputs of the di#erent predictors are non linearly combined through a gating network; ensemble constructed by subsets of input features [10] where each predictor selects a group of the input features. This paper focuses on Error Correcting Output Coding (ECOC) decomposition methods [16, 17, 32, 31, 28] and in particular on the factors a#ecting the e#ectiveness of these ensemble methods. Error correcting output codes [8] originally proposed to enhance the reliability of the transmission of binary signals through a noisy channel [35] have been successfully used in the framework of ....

....codes [52] In fact, using codewords for coding the classes suggest the introduction of codes with error recovering abilities. Kong and Dietterich showed that ECOC techniques can also provide class probability informations, through the solution of an over constrained system of linear equations [31]. An interesting extension of this approach presented by Schapire, consists of the combination of error correcting output codes with boosting techniques [19] this ensemble method shows good performances on di#erent benchmark machine learning problems [49] From a statistical standpoint ECOC ....

B.E. Kong and T.G. Dietterich. Probability estimation via error correcting output coding. In IASTED International Conference: Artificial Intelligence and Soft Computing, Ban#, Canada.

....Codes (ECOC) method being just one of them. In this paper we report on the novel use of ECOC for designing multiple experts for face verification. Use of ECOC for decomposing a multi class problem into a set of complementary two class problems is a well established method in many applications [4, 5, 6, 7, 11, 21, 23, 22, 24]. Such a decomposition means that attention can be focused on developing an effective technique for the two class classifier, without having to consider explicitly the design and automation of the multi class case. It is also hoped that the parameters of a simple expert run many times may be ....

E.B. Kong and T.G. Diettrich. Probability estimation via error-correcting output coding. In Int. Conf. of Artificial Inteligence and soft computing, Banff,Canada, 1997. http://www.cs.orst.edu/ tgd/cv/pubs.html.

....they perform. This is also useful if we want to extend ECOC to deal with applications for which it would be desirable to understand ECOC features as estimation measures. In this paper we look at an alternative ECOC combination strategy based on Least Squares (LS ECOC) which was investigated in [11] and extended by incorporating ridged regression when b is small [8] Recovering individual class probabilities from super class probabilities is easily accomplished by matrix inversion when the individual probability estimates are exact and columns of ECOC matrix are arranged in one per class ....

E.B. Kong and T.G. Diettrich. Probability estimation via error-correcting output coding. In Int. Conf. of Articial Inteligence and soft computing, Ban,Canada, 1997. http://www.cs.orst.edu/ tgd/cv/pubs.html.

No context found.

E.B. Kong and T.G. Diettrich. Probability estimation via error-correcting output coding. Banff,Canada, 1997. Int. Conf. on Artificial Intelligence and soft computing. http://www.cs.orst.edu/ tgd/cv/pubs.html.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC