초록 ▼

A method for and system for training a connection network located between neuron layers within a multi-layer physical neural network. A multi-layer physical neural network can be formed having a plurality of inputs and a plurality outputs thereof, wherein the multi-layer physical neural network comp

A method for and system for training a connection network located between neuron layers within a multi-layer physical neural network. A multi-layer physical neural network can be formed having a plurality of inputs and a plurality outputs thereof, wherein the multi-layer physical neural network comprises a plurality of layers, wherein each layer comprises one or more connection networks and associated neurons. Thereafter, a training wave can be initiated across the connection networks associated with an initial layer of the multi-layer physical neural network which propagates thereafter through succeeding connection networks of succeeding layers of the neural network by successively closing and opening switches associated with each layer. One or more feedback signals thereof can be automatically provided to strengthen or weaken nanoconnections associated with each connection network.

대표청구항 ▼

The embodiments of an invention in which an exclusive property or right is claimed are defined as follows: 1. A method for training at least one connection network located between layers of a multi-layer physical neural network formed utilizing nanotechnology, said method comprising the steps of: f

The embodiments of an invention in which an exclusive property or right is claimed are defined as follows: 1. A method for training at least one connection network located between layers of a multi-layer physical neural network formed utilizing nanotechnology, said method comprising the steps of: forming a multi-layer physical neural network having a plurality of inputs and a plurality outputs thereof, wherein said multi-layer physical neural network comprises a plurality of layers therein, such that each layer thereof comprises at least one connection network, wherein said at least one connection network comprises a plurality of nanoconnections suspended in a solvent; initiating a training wave across said at least one connection network associated with an initial layer of said multi-layer physical neural network which propagates back through preceding connection networks of preceding layers of said multi-layer physical neural network by successively closing at least one switch associated with each layer of said multi-layer physical neural network; and automatically providing at least one feedback signal thereof to each preceding connection network associated with each preceding layer thereof to strengthen or weaken nanoconnections of each connection network of said multi-layer physical neural network. 2. The method of claim 1 further comprising the step of: forming a network structure from said multi-layer physical network, wherein said network structure automatically adapts itself to an input/output relationship thereof without the aid of external network error minimization. 3. The method of claim 1 further comprising the step of: forming a network structure from said multi-layer physical network wherein information thereof is stored and processed by said at least one connection network, including said plurality of nanoconnections suspended in said solvent, thereby absolving a need for separate weight memory storage and processing architectures thereof. 4. The method of claim 1 wherein said solvent comprises at least one of the following types of solvents: a dielectric solvent; liquid crystal media; and a gas. 5. The method of claim 1 wherein each nanoconnection of said plurality of nanoconnections comprises a nanoconductor. 6. The method of claim 1 wherein said nanoconductor comprises a nanotube. 7. The method of claim 1 wherein said nanoconductor comprises a nanowire. 8. The method of claim 1 wherein said nanoconductor comprises a plurality of nanoparticles. 9. The method of claim 1 wherein at least one neuron associated with said connection network comprises an inhibitory neuron. 10. The method of claim 1 wherein at least one neuron associated with said connection network comprises an excitatory neuron. 11. A method for training at least one connection network located between layers of a multi-layer physical neural network formed utilizing nanotechnology, said method comprising the steps of: forming a multi-layer physical neural network having a plurality of inputs and a plurality outputs thereof, wherein said multi-layer physical neural network comprises a plurality of layers therein, such that each layer thereof at least one connection network comprising a plurality of nanoconnections suspended in a dielectric solvent, wherein each nanoconnection of said plurality of nanoconnections comprises a nanoconductor and wherein at least one neuron associated with said connection network comprises at least one inhibitory neuron and at least one excitatory neuron; initiating a training wave across said at least one connection network associated with an initial layer of said multi-layer physical neural network which propagates back through preceding connection networks of preceding layers of said multi-layer physical neural network by successively closing at least one switch associated with each layer of said multi-layer physical neural network; and automatically providing at least one feedback signal thereof to each preceding connection network associated with each preceding layer thereof to strengthen or weaken nanoconnections of each connection network of said multi-layer physical neural network; forming a network structure from said multi-layer physical network, wherein said network structure automatically adapts itself to an input/output relationship thereof without the aid of external network error minimization; and wherein said network structure automatically adapts itself to an input/output relationship thereof without the aid of external digital network error minimization, such that information thereof is stored and processed by said at least one connection network, including said plurality of nanoconnections suspended in said solvent, thereby absolving a need for separate weight memory storage and processing architectures thereof. 12. A system for training at least one connection network located between layers of a multi-layer physical neural network formed utilizing nanotechnology, said system comprising: a multi-layer physical neural network having a plurality of inputs and a plurality outputs thereof, wherein said multi-layer physical neural network comprises a plurality of layers therein, wherein each layer thereof comprises at least one connection network comprising a plurality of nanoconnections suspended in a solvent; and a training wave initiated across said at least one connection network associated with an initial layer of said multi-layer physical neural network, which propagates back through preceding connection networks of preceding layers of said multi-layer physical neural network by successively closing at least one switch associated with each layer of said multi-layer physical neural network; and a feedback mechanism for automatically providing at least one feedback signal provided to each preceding connection network associated with each preceding layer thereof to strengthen or weaken nanoconnections of each connection network of said multi-layer physical neural network. 13. The system of claim 12 further comprising: a network structure formed from said multi-layer physical network, wherein said network structure automatically adapts itself to an input/output relationship thereof without the aid of external network error minimization. 14. The system of claim 12 further comprising: a network structure formed from said multi-layer physical network wherein information thereof is stored and processed by said at least one connection network, including said plurality of nanoconnections suspended in said solvent, thereby absolving a need for separate weight memory storage and processing architectures thereof. 15. The system of claim 12 wherein said solvent comprises at least one of the following: a dielectric solvent; liquid crystal medial; and a gas. 16. The system of claim 12 wherein each nanoconnection of said plurality of nanoconnections comprises a nanoconductor. 17. The system of claim 12 wherein said nanoconductor comprises a nanotube. 18. The system of claim 12 wherein said nanoconductor comprises a nanowire. 19. The system of claim 12 wherein said nanoconductor comprises a plurality of carbon particles. 20. The system of claim 12 wherein at least one neuron associated with said connection network comprises an inhibitory neuron and/or an excitatory neuron.