L2-nonexpansive neural networks
AssigneeInternational Business Machines Corporation
InventorsHaifeng Qian, Mark Wegman
Life of this patent
Estimating the term…
Abstract
A training method, system, and computer program product include training a neural network including at least one of using norm-pooling as a non-linear function, using a two-sided ReLU as a non-linear function, and increasing a confidence gap and further training such that the network comprises a non-expansive network.
Claims20
A computer-implemented training method for an L2 non-expansive neural network, the method comprising:
- training the L2 non-expansive neural network via a multi-sided ReLU as a non-linear function.
A computer program product for training an L2 non-expansive neural network, the computer program product comprising a computer-readable storage medium having program instructions embodied therewith, the program instructions executable by a computer to cause the computer to perform:
- training the L2 non-expansive neural network via a multi-sided ReLU as a non-linear function.
A training system for an L2 non-expansive neural network, the system comprising:
- a processor;
- and
- a memory, the memory storing instructions to cause the processor to perform:
- training the L2 non-expansive neural network via a multi-sided ReLU as a non-linear function.
Dependencies are inferred from the claim text.