Advertisement

Adaptive Resonance Theory Microchips

  • T. Serrano-Gotarredona
  • B. Linares-Barranco
Part of the Studies in Fuzziness and Soft Computing book series (STUDFUZZ, volume 43)

Abstract

This chapter addresses implementation of learning models inspired by Adaptive Resonance Theory. A real-time clustering microchip based on the ART1 algorithm is presented, capable of classification and fast learning of 100-bit input patterns into up to 18 categories. A second chip is also presented which implements the ARTMAP algorithm for pattern association. We include experimental learning results from both prototypes, fabricated in standard single-poly double-metal 1.6μm and 1.0μm CMOS digital processes. MOS Transistors mismatch characterization is­sues are also addressed.

Keywords

Current Source Input Pattern Very Large Scale Integration Current Mirror NMOS Transistor 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. [1]
    Mauduit, N., Duranton, M., Gobert, J., and Sirat, J.A. (1992), “Lneuro 1.0: A piece of hardware lego for building neural network systems,” IEEE Transactions on Neural Networks, Vol. 3, no. 3, pp. 414–422.CrossRefGoogle Scholar
  2. [2]
    Jones, S., Sammut, K., Nielsen, C, and J. Staunstrup, (1991), “Toroidal neural network: Architecture and processor granularity issues,” in VLSI Design of Neural Networks, U. Ramacher and U. Rueckert, Eds., pp. 229–254. Kluwer Academic Publishers, Dor­drecht, Netherlands.CrossRefGoogle Scholar
  3. [3]
    Ramacher, U., Beichter, J., Raab, W., Anlauf, J., Bruels, N., Hachmann, U., and Wesseling, M. (1991), “Design of a 1st generation neurocomputer,” in VLSI Design of Neural Networks, Ramacher, U. and Rueckert, U. Eds., pp. 271–310. Kluwer Academic Publishers, Dordrecht, Netherlands.CrossRefGoogle Scholar
  4. [4]
    Kohonen, T. (1989), Self-Organization and Associative Memory, Springer-Verlag, Berlin, Germany.CrossRefGoogle Scholar
  5. [5]
    Bezdek, J. (1981), Pattern Recognition with Fuzzy Objective Function Algorithms, Plenum, New York.MATHCrossRefGoogle Scholar
  6. [6]
    Duda, R. and Hart, P. (1973), Pattern Classification and Scene Analysis, Wiley, New York.MATHGoogle Scholar
  7. [7]
    Hartigan, J. (1975), Clustering Algorithms, Wiley, New York.MATHGoogle Scholar
  8. [8]
    Dubes, R. and Jain, A. (1988), Algorithms that Cluster Data, Pren­tice Hall, Englewood Cliffs, NJ.Google Scholar
  9. [9]
    Pao, Y.H. (1989), Adaptive Recognition and Neural Networks, Addison Wesley, Reading, MA.MATHGoogle Scholar
  10. [10]
    Rodriguez-Vazquez, A., Espejo, S., Dommguez-Castro, R., Huertas, J.L., and Sanchez-Sinencio, E. (1993), “Current-mode techniques for the implementation of continous- and discrete-time cellular neural networks,” IEEE Transactions on Circuits and Systems II, Vol 40, no. 3, pp. 132–146.CrossRefGoogle Scholar
  11. [11]
    Linares-Barranco, B., Sanchez-Sinencio, E., Rodriguez-Vazquez, A., and Huertas, J.L. (1993), “A CMOS analog adaptive BAM with on-chip learning and weight refreshing,” IEEE Transactions on Neural Networks, Vol. 4, no. 3, pp. 445–455.CrossRefGoogle Scholar
  12. [12]
    Carpenter, G.A. and Grossberg, S. (1987), “A massively parallel architecture for a self-organizing neural pattern recognition machine,” Computer Vision, Graphics, and Image Processing, Vol. 37, pp. 54–115.MATHCrossRefGoogle Scholar
  13. [13]
    Serrano-Gotarredona, T. and Linares-Barranco, B. (1998), Adaptive Resonance Theory Microchips. Circuit Design Techniques, Kluwer Academic Publishers, Boston.CrossRefGoogle Scholar
  14. [14]
    Keulen, E., Colak, S., Withagen, H., and Hegt, H. (1994), “Neural network hardware performance criteria,” Proceedings of the 1994 Int. Conf. on Neural Networks (ICNNf 94), pp. 1885–1888.Google Scholar
  15. [15]
    Riedmiller, M. (1994), “Advanced supervised learning in multi­layer perceptions — from backpropagation to adaptive learning algorithms,” Int. Journal of Computer Standards and Interfaces, Vol. 5.Google Scholar
  16. [16]
    Kosko, B. (1987), “Adaptive bidirectional associative memories,” Applied Optics, Vol. 26, pp. 4947–4960.CrossRefGoogle Scholar
  17. [17]
    Wand, Y.F., Cruz, J.B., and Mulligan, J.H. (1990), “On multiple training for bidirectional associative memory,” IEEE Transactions on Neural Networks, Vol. 1, pp. 275–276.CrossRefGoogle Scholar
  18. [18]
    Linares-Barranco, B., Sanchez-Sinencio, E., Rodriguez-Vazquez, A., and Huertas, J.L. (1992), “A modular T-mode design approach for analog neural network hardware implementations, ” IEEE Journal of Solid-State Circuits, Vol. 27, no. 5, pp. 701–713.CrossRefGoogle Scholar
  19. [19]
    Ho, C.S., Liou, JJ., Georgiopoulos, M., Heileman, G.L., and Christodoulo, C. (1994), “Analogue circuit design and implemen­tation of an adaptive resonance theory (ART) neural network ar­chitecture, ” International Journal of Electronics, Vol. 76, no. 2, pp. 271–291.CrossRefGoogle Scholar
  20. [20]
    Tsay, S.W. and Newcomb, R.W. (1991), “VLSI implementation of ART1 memories, ” IEEE Transactions on Neural Networks, Vol. 2, pp. 214–221, March.CrossRefGoogle Scholar
  21. [21]
    Wunsch II, D.C., Caudell, T.P., Capps, CD., Marks II, R.J., and Falk, R.A. (1993), “An optoelectronic implementation of the adaptive resonance neural network, ” IEEE Transactions on Neural Networks, Vol. 4, no. 4, pp. 673–684, July.CrossRefGoogle Scholar
  22. [22]
    Serrano-Gotarredona, T. and Linares-Barranco, B. (1996), “A modified ART1 algorithm more suitable for VLSI implementations, ” Neural Networks, Vol. 9, no. 6, pp. 1025–1043.CrossRefGoogle Scholar
  23. [23]
    Serrano-Gotarredona, T. and Linares-Barranco, B. (1996), “A realtime clustering microchip neural engine, ” IEEE Transactions on Very Large Scale Integration (VLSI) Systems, Vol. 4, no. 2, pp. 195–209.CrossRefGoogle Scholar
  24. [24]
    Lazzaro, J., Ryckebush, R., Mahowald, M.A., and Mead, C (1989), “Winner-take-all networks of O(n) complexity, ” in Advances in Neural Information Processing Systems, Touretzky, D.S. Ed., Vol. 1, pp. 703–711. Morgan-Kaufmann, Los Altos, CA.Google Scholar
  25. [25]
    Rodriguez-Vazquez, A., Dominguez-Castro, R., Medeiro, R, and Delgado-Resti-tuto, M. (1995), “High resolution CMOS current comparators: Design and application to current-mode function generation,” International Journal of Analog Integrated Circuits and Signal Processing, Vol. 7, pp. 149–165.CrossRefGoogle Scholar
  26. [26]
    Mead, C. (1989), Analog VLSI and Neural Systems, Addison Wesley, Reading, MA.MATHCrossRefGoogle Scholar
  27. [27]
    Serrano-Gotarredona, T. and Linares-Barranco, B. (1994), “The active-input regulated-cascode current mirror, ” IEEE Transactionson Circuits and Systems, Parti: Fundamental Theory and Applications, Vol. 41, no. 6, pp. 464–467.CrossRefGoogle Scholar
  28. [28]
    Serrano-Gotarredona, T. and Linares-Barranco, B. (1995), “A modular current-mode high-precision winner-take-all circuit,” IEEETransactions on Circuits and Systems, Part II, Vol. 42, no. 2, pp. 132–134.CrossRefGoogle Scholar
  29. [29]
    Serrano-Gotarredona, T. and Linares-Barranco, B. (1998), “A high-precision current-mode WTA-MAX circuit with multi-chip capability,” IEEE Journal of Solid-State Circuits, Vol. 33, no. 2, pp. 280–286.CrossRefGoogle Scholar
  30. [30]
    Pelgrom, M.J.M., Duinmaijer, A.C.J., and Welbers, A.P.G. (1989), “Matching properties of MOS transistors,” IEEE Journal ofSolid-State Circuits, Vol. 24, pp. 1433–1440.CrossRefGoogle Scholar
  31. [31]
    Serrano-Gotarredona, T. and Linares-Barranco, B. (1999), “Sys­tematic width-and-length dependent cmos transistor mismatch characterization and simulation,” International Journal of Analog Integrated Circuits and Signal Processing. Google Scholar
  32. [32]
    Gregor, R.W. (1992), “On the relationship between topography and transistor matching in an analog CMOS technology,” IEEE Transactions on Electron Devices, Vol. 39, no. 2, pp. 275–282.CrossRefGoogle Scholar
  33. [33]
    Lakshmikumar, K.R., Hadaway, R.A., and Copeland, M.A. (1986), “Characterization and modeling of mismatch in MOS transistors for precision analog design,” IEEE Journal of Solid-State Circuits, Vol. SC-21, no. 6, pp. 1057–1066.CrossRefGoogle Scholar
  34. [34]
    Michael, C. and Ismail, M. (1992), “Statistical modeling of device mismatch for analog MOS integrated circuits,” IEEE Journal of Solid-State Circuits, Vol. 27, no. 2, pp. 154–166.CrossRefGoogle Scholar
  35. [35]
    Shyu, J.B., Temes, G.C., and Krummenacher, F. (1984), “Random error effects in matched MOS capacitors and current sources,” IEEE Journal of Solid-State Circuits, Vol. SC-19, no. 6, pp. 948–955.CrossRefGoogle Scholar
  36. [36]
    Allen, RE. and Holberg, D.R. (1987), CMOS Analog Design, Holt Rinehart and Winston Inc., New York.Google Scholar
  37. [37]
    Strader, N.R. and Harden, J.C. (1989), “Architectural yield optimization,” in Wafer Scale Integration, Schwartzlander, E.E. Ed., pp. 57–118. Kluwer Academic Publishers, Boston.CrossRefGoogle Scholar
  38. [38]
    Chu, L.C. (1991), “Fault-tolerant model of neural computing,” IEEE Int. Conf on Computer Design: VLSI in Computers and Pro-cessors, pp. 122–125.Google Scholar
  39. [39]
    Neti, C, Schneider, M.H., and Young, E.D. (1992), “Maximally fault tolerant neural networks,” IEEE Transactions on Neural Networks, Vol. 3, no. 1, pp. 14–23.CrossRefGoogle Scholar
  40. [40]
    Kim, J.H., Lursinsap, C, and Park, S. (1991), “Fault-tolerant ar­tificial neural networks,” Int. Joint Conf. on Neural Networks (IJCNNy 91), Vol. 2, pp. 951.CrossRefGoogle Scholar
  41. [41]
    Serrano-Gotarredona, T. and Linares-Barranco, B. (1997), “An ART1 microchip and its use in multi-ARTl systems,” IEEE Trans­actions on Neural Networks, Vol. 8, no. 5, pp. 1184–1194.CrossRefGoogle Scholar
  42. [42]
    Carpenter, G.A., Grossberg, S., and Reynolds, J.H. (1991), “ARTMAP: Supervised real-time learning and classification of nonstationary data by a self-organizing neural network,” NeuralNetworks, Vol. 4, pp. 759–771.Google Scholar
  43. [43]
    Medeiro, E, Rodriguez-Macias, R., Fernandez, F.V., Domfnguez-Castro, R., and Rodriguez-Vazquez, A. (1994), “Global analogue cell design and optimization using statistical techniques,” Analog Integrated Circuits and Signal Processing, Vol. 6, pp. 179–195.CrossRefGoogle Scholar
  44. [44]
    Bastos, J., Steyaert, M., Roovers, R., Kinger, P., Sansen, W., Graindourze, B., Pergoot, A., and Janssens, E. (1995), “Mismatch char­acterization of small size MOS transistors” Proc. IEEE 1995 Int. Conf. Microelectronic Test Structures, Vol. 8, pp. 271–276.Google Scholar
  45. [45]
    Bastos, J., Steyaert, M., Pergoot, A., and Sansen, W. (1997), “Mismatch characterization of submicron MOS transistors,” Analog Integrated Circuits and Signal Processing, Vol. 12, pp. 95–106.CrossRefGoogle Scholar
  46. [46]
    Rade, L. and Westergren, B. (1990), BETA Mathematics Handbook, CRC Press.MATHGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2000

Authors and Affiliations

  • T. Serrano-Gotarredona
    • 1
  • B. Linares-Barranco
    • 1
  1. 1.National Microelectronics CenterSevillaSpain

Personalised recommendations