Methods and Systems for Detecting Spurious Data Patterns

1 . A method for detection and classification of data, the method comprising:

receiving input data at a neural network circuit comprising a plurality of node layers, with each of the plurality of node layers comprising respective one or more nodes, the neural network circuit further comprising adjustable weighted connections connecting at least some nodes in different layers of the plurality of node layers; and

removing one or more of the weighted connections at one or more time instances.

2 . The method of claim 1 , wherein the neural network circuit is a feed-forward neural network circuit.

3 . The method of claim 1 , wherein removing the one or more of the weighted connections comprises:

selecting the one or more of the weighted connections randomly; and

removing the randomly selected one or more of the weighted connections.

4 . The method of claim 1 , wherein removing the one or more of the weighted connections comprises:

selecting a set of multiple connections from the weighted connections based, at least in part, on output of the neural network circuit; and

selecting randomly the one or more of the weighted connections from the selected set of multiple connections.

5 . The method of claim 4 , wherein selecting the set of multiple connections comprises:

selecting one or more pairs of node layers of the neural network circuit according to the output of the neural network circuit; and

removing at least one weighted connection between node layers of the selected one or more pairs of node layers.

6 . The method of claim 4 , wherein selecting the set of multiple connections comprises:

selecting the set of multiple connections according to output values produced by elements of an output node layer of the neural network circuit and a plurality of output ranges defined for possible values produced by the output node layer.

7 . The method of claim 1 , further comprising:

configuring at least some of the weighted connections according to a biasing factor in response to output of the neural network resulting from an input data record, of the received input data, processed by the neural network.

8 . The method of claim 7 , wherein the biasing factor is a multiplication factor applied to the at least some of the weighted connections through a back-propagation operation in response to a determination that the neural network correctly identified the input data record as being anomalous.

9 . The method of claim 1 , further comprising:

performing preprocessing on a received raw data record to produce an input data record provided to the neural network circuit, including performing one or more of: Gaussian normalization applied to the raw data record, or removing one or more data elements of the raw data record based on at least one of: entropy associated with the one or more data elements, sparseness associated with the one or more data elements, a p-value associated with the one or more data elements, or a low-effect value associated with the one or more data elements.

10 . The method of claim 9 , wherein removing one or more data elements comprises:

identifying a particular data element as a rare element in response to determining, based on training data to train a learning engine implementation for performing the preprocessing, that the particular data element is present in fewer than an adjustable threshold number of data records comprising the training data, wherein the adjustable threshold number is adjusted based on likelihood of occurrence of anomalous values for the particular data element; and

removing from runtime data records the particular data element identified as the rare element.

11 . A system comprising:

an input stage to receive one or more input data records; and

a controller, implementing one or more learning engines, in communication with a memory device to store programmable instructions, to:

receive input data at a neural network circuit comprising a plurality of node layers, with each of the plurality of node layers comprising respective one or more nodes, the neural network circuit further comprising adjustable weighted connections connecting at least some nodes in different layers of the plurality of node layers; and

remove one or more of the weighted connections at one or more time instances.

12 . The system of claim 11 , wherein the neural network circuit includes a feed-forward neural network circuit.

13 . The system of claim 11 , wherein the controller configured to remove the one or more of the weighted connections is configured to:

select the one or more of the weighted connections randomly; and

remove the randomly selected one or more of the weighted connections.

14 . The system of claim 11 , wherein the controller configured to remove the one or more of the weighted connections is configured to:

select a set of multiple connections from the weighted connections based, at least in part, on output of the neural network circuit; and

select randomly the one or more of the weighted connections from the selected set of multiple connections.

15 . The system of claim 14 , wherein the controller configured to select the set of multiple connections is configured to:

select one or more pairs of node layers of the neural network circuit according to the output of the neural network circuit; and

remove at least one weighted connection between node layers of the selected one or more pairs of node layers.

16 . The system of claim 14 , wherein the controller configured to select the set of multiple connections is configured to:

select the set of multiple connections according to output values produced by elements of an output node layer of the neural network circuit and a plurality of output ranges defined for possible values produced by the output node layer.

17 . The system of claim 11 , wherein the controller is further configured to:

configure at least some of the weighted connections according to a biasing factor in response to output of the neural network resulting from an input data record, of the received input data, processed by the neural network.

18 . The system of claim 17 , wherein the biasing factor is a multiplication factor applied to the at least some of the weighted connections through a back-propagation operation in response to a determination that the neural network correctly identified the input data record as being anomalous.

19 . The system of claim 11 , wherein the controller is further configured to:

perform preprocessing on a received raw data record to produce an input data record provided to the neural network circuit, including performing one or more of: Gaussian normalization applied to the raw data record, or removing one or more data elements of the raw data record based on at least one of: entropy associated with the one or more data elements, sparseness associated with the one or more data elements, a p-value associated with the one or more data elements, or a low-effect value associated with the one or more data elements.

20 . A non-transitory computer readable media storing a set of instructions, executable on at least one programmable device, to:

receive input data at a neural network circuit comprising a plurality of node layers, with each of the plurality of node layers comprising respective one or more nodes, the neural network circuit further comprising adjustable weighted connections connecting at least some nodes in different layers of the plurality of node layers; and

remove one or more of the weighted connections at one or more time instances.