Minimal Learning Machine: A New Distance-Based Method for Supervised Learning

详细信息    查看全文

• 作者：Amauri Holanda de Souza Junior (19)
  Francesco Corona (21)
  Yoan Miche (21)
  Amaury Lendasse (21)
  Guilherme A. Barreto (20)
  Olli Simula (21)
  
• 刊名：Lecture Notes in Computer Science
• 出版年：2013
• 出版时间：2013
• 年：2013
• 卷：7902
• 期：1
• 全文大小：501KB
• 参考文献：1. Cox, T., Cox, M.: Multidimensional Scaling. Chapman & Hall, London (1994)
  2. de Silva, V., Tenenbaum, J.B.: Global versus local methods in nonlinear dimensionality reduction. In: Advances in Neural Information Processing Systems, vol.聽15, pp. 705鈥?12 (2002)
  3. Huang, G.B., Zhu, Q.Y., Ziew, C.K.: Extreme Learning Machine: Theory and applications. Neurocomputing聽70(1-3), 489鈥?01 (2006) CrossRef
  4. Miche, Y., Sorjamaa, A., Bas, P., Simula, O., Jutten, C., Lendasse, A.: OP-ELM: Optimally Pruned Extreme Learning Machine. IEEE Transactions on Neural Networks聽21(1), 158鈥?62 (2010) CrossRef
  5. Smola, A.J., Sch枚lkopf, B.: A tutorial on support vector regression. Statistics and Computing聽14(3), 199鈥?22 (2004) CrossRef
  6. Rasmussen, C.E., Williams, C.K.I.: Gaussian Processes for Machine Learning. The MIT Press, Cambridge (2006)
  7. Bishop, C.M.: Neural Networks for Pattern Recognition. Oxford University Press, New York (1995)
  8. Chang, C.-C., Lin, C.-J.: LIBSVM: A Library for Support Vector Machines. ACM Transactions on Intelligent Systems and Technology聽2(27), 1鈥?7 (2011) CrossRef
  9. Lehmann, E.L., Romano, J.P.: Testing statistical hypotheses. Springer, New York (2005)
  
• 作者单位：Amauri Holanda de Souza Junior (19)
  Francesco Corona (21)
  Yoan Miche (21)
  Amaury Lendasse (21)
  Guilherme A. Barreto (20)
  Olli Simula (21)
  
  19. Department of Computer Science, Federal Institute of Cear谩, Av. Contorno Norte, 10, Maracana煤, Cear谩, Brazil
  21. Department of Information and Computer Science, Aalto University, Konemiehentie 2, Espoo, Finland
  20. Department of Teleinformatics Engineering, Federal University of Cear谩, Av. Mister Hull, S/N - Campus of Pici, Center of Technology, Fortaleza, Cear谩, Brazil
  
• ISSN：1611-3349

文摘

In this work, a novel supervised learning method, the Minimal Learning Machine (MLM), is proposed. Learning a MLM consists in reconstructing the mapping existing between input and output distance matrices and then estimating the response from the geometrical configuration of the output points. Given its general formulation, the Minimal Learning Machine is inherently capable to operate on nonlinear regression problems as well as on multidimensional response spaces. In addition, an intuitive extension of the MLM is proposed to deal with classification problems. On the basis of our experiments, the Minimal Learning Machine is able to achieve accuracies that are comparable to many de facto standard methods for regression and it offers a computationally valid alternative to such approaches.