Large Scale Indefinite Kernel Fisher Discriminant
Indefinite similarity measures can be frequently found in bio-informatics by means of alignment scores. Lacking an underlying vector space, the data are given as pairwise similarities only. Indefinite Kernel Fisher Discriminant (iKFD) is a very effective classifier for this type of data but has cubic complexity and does not scale to larger problems. Here we propose an extension of iKFD such that linear runtime and memory complexity is achieved for low rank indefinite kernels. Evaluation at several larger similarity data from various domains shows that the proposed method provides similar generalization capabilities while being substantially faster for large scale data.
A Marie Curie Intra-European Fellowship (IEF): FP7-PEOPLE-2012-IEF (FP7-327791-ProMoS) and support from the Cluster of Excellence 277 Cognitive Interaction Technology funded by the German Excellence Initiative is gratefully acknowledged. PT was supported by the EPSRC grant EP/L000296/1, “Personalized Health Care through Learning in the Model Space”. We would like to thank R. Duin, Delft University for various support with distools and prtools and Huanhuan Chen,University of Science and Technology of China, for providing support with the Probabilistic Classification Vector Machine.
- 1.Alabdulmohsin, I.M., Gao, X., Zhang, X.: Support vector machines with indefinite kernels. In: Phung, D., Li, H. (eds.) Proceedings of the Sixth Asian Conference on Machine Learning, ACML 2014. JMLR Proceedings, Nha Trang City, Vietnam, 26–28 November 2014, vol. 39 (2014). JMLR.org
- 4.Chen, H., Tino, P., Yao, X.: Efficient probabilistic classification vector machine with incremental basis function selection. IEEE TNN-LS 25(2), 356–369 (2014)Google Scholar
- 6.Diethe, T., Hussain, Z., Hardoon, D.R., Shawe-Taylor, J.: Matching pursuit kernel fisher discriminant analysis. In: Dyk, D.A.V., Welling, M. (eds.) Proceedings of the Twelfth International Conference on Artificial Intelligence and Statistics, AISTATS 2009. JMLR Proceedings, 16–18 April 2009, Clearwater Beach, Florida, USA, vol. 5, pp. 121–128 (2009). JMLR.org
- 7.Dubuisson, M., Jain, A.: A modified hausdorff distance for object matching. In: Proceedings of the 12th IAPR International Conference on Pattern Recognition, 1994. Vol. 1 - Conference A: Computer Vision amp; Image Processing., vol. 1, pp. 566–568, October 1994Google Scholar
- 8.Duin, R.P.: prtools, March 2012Google Scholar
- 9.Gisbrecht, A., Schleif, F.-M.: Metric and non-metric proximity transformations at linear costs. Neurocomputing (2015, to appear)Google Scholar
- 11.Haasdonk, B., Pekalska, E.: Indefinite kernel fisher discriminant. In: 19th International Conference on Pattern Recognition (ICPR 2008), 8–11 December 2008, Tampa, Florida, USA, pp. 1–4. IEEE Computer Society (2008)Google Scholar
- 16.Schleif, F.-M., Gisbrecht, A., Tino, P.: Probabilistic classification vector machine at large scale. In: Proceedings of ESANN 2015, pp. 555–560 (2015)Google Scholar
- 18.Schleif, F.-M., Tino, P.: Indefinite proximity learning - a review. Neural Computation (2015, to appear)Google Scholar
- 20.Williams, C.K.I., Seeger, M.: Using the nyström method to speed up kernel machines. In: NIPS 2000, pp. 682–688 (2000)Google Scholar