Thresholded Learning Matrix for Efficient Pattern Recalling
The Lernmatrix, which is the first known model of associative memory, is a heteroassociative memory that can easily work as a binary pattern classifier if output patterns are appropriately chosen. However, this mathematical model undergoes fundamental patterns misclassification whenever crossbars saturation occurs. In this paper, a novel algorithm that overcomes Lernmatrix weaknesses is proposed. The crossbars saturation occurrence is solved by means of a dynamic threshold value which is computed for each recalled pattern. The algorithm applies the dynamic threshold value over the ambiguously recalled class vector in order to obtain a sentinel vector which is used for uncertainty elimination purposes. The efficiency and effectiveness of our approach is demonstrated through comparisons with other methods using real-world data.
KeywordsAssociative Memories Dynamic Threshold Lernmatrix Pattern Classification Supervised Learning
- 4.Yáñez-Márquez, C., Díaz-de-León Santiago, J.L.: Lernmatrix de Steinbuch. IT-48 Serie Verde, CIC-IPN, México (2001)Google Scholar
- 6.Yáñez-Márquez, C.: Associative Memories Based on Order Relations and Binary Operators (in Spanish). PhD Thesis. Center for Computing Research, México (2002)Google Scholar
- 7.Chren, W.A.: One-hot residue coding for high-speed non-uniform pseudo-random test pattern generation. In: IEEE International Symposium on Circuits and Systems ISCAS 1995, pp. 401–404. IEEE Press, New York (1995)Google Scholar
- 8.Sánchez Garfias, F.A., Díaz-de-León Santiago, J.L., Yáñez Márquez, C.: Lernmatrix de Steinbuch: condiciones necesarias y suficientes para recuperación perfecta de patrones. Research on Computing Science 1, 437–448 (2002)Google Scholar
- 9.Sánchez Garfias, F.A., Díaz-de-León Santiago, J.L., Yáñez Márquez, C.: New Results on the Lernmatrix Properties. Research on Computing Science 10, 91–102 (2004)Google Scholar