An experimental analog CMOS self-learning chip

The analog VLSI implementation of an on-chip learning neural network is discussed in this paper. The multi layer perceptron paradigm and back propagation learning rule have been mapped onto analog circuits. A local learning rate adaptation rule has been also considered to improve the training performance (i.e., fast convergence speed). Experimental results confirm the chip functionality and the soundness of our approach.

[1]  Tor Sverre Lande,et al.  An analog feed-forward neural network with on-chip learning , 1996 .

[2]  Takashi Morie,et al.  An all-analog expandable neural network LSI with on-chip backpropagation learning , 1994, IEEE J. Solid State Circuits.

[3]  Carver Mead,et al.  Analog VLSI and neural systems , 1989 .

[4]  G. M. Bo,et al.  Modeling of current switching during weight update in the analog VLSI implementation of the back-propagation algorithm , 1997 .

[5]  Daniele D. Caviglia,et al.  A Circuit Architecture for Analog On-Chip Back Propagation Learning with Local Learning Rate Adaptation , 1999 .

[6]  Anders Krogh,et al.  Introduction to the theory of neural computation , 1994, The advanced book program.

[7]  Eric A. Vittoz,et al.  Analog VLSI signal processing: Why, where, and how? , 1994 .

[8]  Anne-Johan Annema.  Feed-forward neural networks , 1995 .

[9]  Gert Cauwenberghs,et al.  An analog VLSI recurrent neural network learning a continuous-time trajectory , 1996, IEEE Trans. Neural Networks.

[10]  Daniele D. Caviglia,et al.  An experimental analog VLSI neural network with on-chip back-propagation learning , 1992, ESSCIRC '92: Eighteenth European Solid-State Circuits conference.

[11]  Alan Kramer.  Array-based analog computation: principles, advantages and limitations , 1996, Proceedings of Fifth International Conference on Microelectronics for Neural Networks.

[12]  Howard C. Card,et al.  Tolerance to analog hardware of on-chip learning in backpropagation networks , 1995, IEEE Trans. Neural Networks.