Analog VLSI and neural systems
Analog VLSI and neural systems
A menu of designs for reinforcement learning over time
Neural networks for control
The computational brain
Technical Note: \cal Q-Learning
Machine Learning
Mean-field theory for batched TD (&lgr;)
Neural Computation
Learning to Predict by the Methods of Temporal Differences
Machine Learning
Summed Weight Neuron Perturbation: An O(N) Improvement Over Weight Perturbation
Advances in Neural Information Processing Systems 5, [NIPS Conference]
A Fast Stochastic Error-Descent Algorithm for Supervised Learning and Optimization
Advances in Neural Information Processing Systems 5, [NIPS Conference]
Analog VLSI Implementation of Gradient Descent
Advances in Neural Information Processing Systems 5, [NIPS Conference]
A Parallel Gradient Descent Method for Learning in Analog VLSI Neural Networks
Advances in Neural Information Processing Systems 5, [NIPS Conference]
An analog VLSI recurrent neural network learning a continuous-time trajectory
IEEE Transactions on Neural Networks
A Nonlinear Noise-Shaping Delta-Sigma Modulator with On-Chip Reinforcement Learning^{*}
Analog Integrated Circuits and Signal Processing - Special issue on Learning on Silicon
VLSI Implementation of Fuzzy Adaptive Resonance and Learning Vector Quantization
Analog Integrated Circuits and Signal Processing
VLSI Implementation of Fuzzy Adaptive Resonance and Learning Vector Quantization
MICRONEURO '99 Proceedings of the 7th International Conference on Microelectronics for Neural, Fuzzy and Bio-Inspired Systems
High-speed, model-free adaptive control using parallel synchronous detection
Proceedings of the 20th annual conference on Integrated circuits and systems design
A SiGe BiCMOS eight-channel multidithering sub-microsecond adaptive controller
IEEE Transactions on Circuits and Systems Part I: Regular Papers
| Hi-index | 0.00 |
We present analog VLSI neuromorphic architectures fora general class of learning tasks, which include supervised learning,reinforcement learning, and temporal difference learning. Thepresented architectures are parallel, cellular, sparse in globalinterconnects, distributed in representation, and robust to noiseand mismatches in the implementation. They use a parallel stochasticperturbation technique to estimate the effect of weight changeson network outputs, rather than calculating derivatives basedon a model of the network. This ’’model-free‘‘ technique avoidserrors due to mismatches in the physical implementation of thenetwork, and more generally allows to train networks of whichthe exact characteristics and structure are not known. With additionalmechanisms of reinforcement learning, networks of fairly generalstructure are trained effectively from an arbitrarily suppliedreward signal. No prior assumptions are required on the structureof the network nor on the specifics of the desired network response.