{"page":"145 - 153","publication":"ICML'13 Proceedings of the 30th International Conference on International","author":[{"id":"2CCAC26C-F248-11E8-B48F-1D18A9856A87","full_name":"Takhanov, Rustem","first_name":"Rustem","last_name":"Takhanov"},{"last_name":"Kolmogorov","first_name":"Vladimir","full_name":"Kolmogorov, Vladimir","id":"3D50B0BA-F248-11E8-B48F-1D18A9856A87"}],"language":[{"iso":"eng"}],"_id":"2272","citation":{"mla":"Takhanov, Rustem, and Vladimir Kolmogorov. “Inference Algorithms for Pattern-Based CRFs on Sequence Data.” <i>ICML’13 Proceedings of the 30th International Conference on International</i>, vol. 28, no. 3, ML Research Press, 2013, pp. 145–53.","ieee":"R. Takhanov and V. Kolmogorov, “Inference algorithms for pattern-based CRFs on sequence data,” in <i>ICML’13 Proceedings of the 30th International Conference on International</i>, Atlanta, GA, USA, 2013, vol. 28, no. 3, pp. 145–153.","chicago":"Takhanov, Rustem, and Vladimir Kolmogorov. “Inference Algorithms for Pattern-Based CRFs on Sequence Data.” In <i>ICML’13 Proceedings of the 30th International Conference on International</i>, 28:145–53. ML Research Press, 2013.","apa":"Takhanov, R., &#38; Kolmogorov, V. (2013). Inference algorithms for pattern-based CRFs on sequence data. In <i>ICML’13 Proceedings of the 30th International Conference on International</i> (Vol. 28, pp. 145–153). Atlanta, GA, USA: ML Research Press.","ista":"Takhanov R, Kolmogorov V. 2013. Inference algorithms for pattern-based CRFs on sequence data. ICML’13 Proceedings of the 30th International Conference on International. ICML: International Conference on Machine Learning, JMLR, vol. 28, 145–153.","ama":"Takhanov R, Kolmogorov V. Inference algorithms for pattern-based CRFs on sequence data. In: <i>ICML’13 Proceedings of the 30th International Conference on International</i>. Vol 28. ML Research Press; 2013:145-153.","short":"R. Takhanov, V. Kolmogorov, in:, ICML’13 Proceedings of the 30th International Conference on International, ML Research Press, 2013, pp. 145–153."},"conference":{"end_date":"2013-06-21","location":"Atlanta, GA, USA","name":"ICML: International Conference on Machine Learning","start_date":"2013-06-16"},"year":"2013","intvolume":"        28","volume":28,"publist_id":"4672","month":"06","article_processing_charge":"No","date_updated":"2023-10-17T09:51:32Z","type":"conference","day":"01","department":[{"_id":"VlKo"}],"main_file_link":[{"url":"http://proceedings.mlr.press/v28/takhanov13.pdf?CFID=105472548&CFTOKEN=5c5859b5d97b4439-27B4AC58-BA92-A964-B598CAACEE6CC515","open_access":"1"}],"abstract":[{"text":"We consider Conditional Random Fields (CRFs) with pattern-based potentials defined on a chain. In this model the energy of a string (labeling) x1...xn is the sum of terms over intervals [i,j] where each term is non-zero only if the substring xi...xj equals a prespecified pattern α. Such CRFs can be naturally applied to many sequence tagging problems.\r\nWe present efficient algorithms for the three standard inference tasks in a CRF, namely computing (i) the partition function, (ii) marginals, and (iii) computing the MAP. Their complexities are respectively O(nL), O(nLℓmax) and O(nLmin{|D|,log(ℓmax+1)}) where L is the combined length of input patterns, ℓmax is the maximum length of a pattern, and D is the input alphabet. This improves on the previous algorithms of (Ye et al., 2009) whose complexities are respectively O(nL|D|), O(n|Γ|L2ℓ2max) and O(nL|D|), where |Γ| is the number of input patterns.\r\nIn addition, we give an efficient algorithm for sampling. Finally, we consider the case of non-positive weights. (Komodakis &amp; Paragios, 2009) gave an O(nL) algorithm for computing the MAP. We present a modification that has the same worst-case complexity but can beat it in the best case. ","lang":"eng"}],"date_published":"2013-06-01T00:00:00Z","publication_status":"published","issue":"3","publisher":"ML Research Press","scopus_import":"1","oa":1,"oa_version":"Submitted Version","date_created":"2018-12-11T11:56:41Z","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","alternative_title":["JMLR"],"title":"Inference algorithms for pattern-based CRFs on sequence data","quality_controlled":"1","related_material":{"record":[{"status":"public","id":"1794","relation":"later_version"}]},"status":"public"}