Danielle Hinton

Learn More
A fast stochastic error-descent algorithm for supervised learning and optimization. (1991). Weight perturbation: An optimal architecture and learning technique for analog vlsi feedforward and recurrent multilayer networks. (1991). Learning continuous probability distributions with the con-trastive Hebbian algorithm. 10 appears in many theories of(More)
1. Note that the maximum number of errors for the classification networks is 504. 1. Note that the maximum number of errors for 'Balance' and 'Depth' is 540, and 504 for the other simulations (since the balanced trees are not included then) especially hard for constituents that appear at deeper levels of the tree, and is also true even if the RAAM can(More)
LIBRARIES 1 Acknowledgments I would like to thank everyone who helped me write this thesis. In Particular, Professor Vahid Tarokh for his invaluable support and guidance. His counsel and deep insight into issues was very enlightening and inspiring. Carol Fredricks and Kate Baty who offered support and encouragement during my research. I would like to thank(More)
  • 1