Statistics-based anomaly detection
First Claim
1. A method for identifying an anomaly in a signal, comprising:
- receiving a discrete signal, having sample values corresponding to amounts of data flow in a network within a time interval;
generating a sequence of likelihoods corresponding to sample values in the signal and based at least in part on a historical probability distribution of previously received sample values corresponding to amounts of data flow in the network, wherein a likelihood is a probability of occurrence of a corresponding sample value in the signal;
identifying likelihood change points in the likelihood sequence by;
selecting a parameter L corresponding to a minimum number of samples in a segment;
appending L consecutive likelihoods to a buffer;
computing a sequence of first sum values of the likelihoods in the buffer;
obtaining a sequence of second sum values;
determining the presence of a change point in the buffer based at least in part on a comparison between the first and second sum values, wherein a plurality of likelihoods preceding the change point have a first statistic value and a plurality of likelihoods following the change point have a second statistic value different from the first statistic value; and
identifying a likelihood in the buffer as a change point based at least in part on the comparison;
segmenting the discrete signal into a plurality of segments at samples corresponding to the identified change points such that a respective one of the samples corresponding to the identified change points is at one of a beginning or an end of each of the plurality of segments;
identifying a segment as an anomaly based on a comparison between a statistic of the segment and a statistic of the historical probability distribution; and
reconfiguring, responsive to identifying the segment as the anomaly, a component of the network.
2 Assignments
0 Petitions
Accused Products
Abstract
Systems and methods are described herein for detecting an anomaly in a discrete signal, where samples in the signal correspond to amounts of data flow in a network within a time interval. The discrete signal is received, and a sequence of likelihoods corresponding to the sample values in the signal is generated. The likelihoods are based at least in part on a historical probability distribution of previously received sample values, and a likelihood is a probability of occurrence of a corresponding sample value in the signal. Likelihood change points are identified in the likelihood sequence, and the discrete signal is segmented into a plurality of segments at samples corresponding to the identified change points. A segment is identified as an anomaly based on a comparison between a statistic of the segment and a statistic of the historical probability distribution.
-
Citations
18 Claims
-
1. A method for identifying an anomaly in a signal, comprising:
-
receiving a discrete signal, having sample values corresponding to amounts of data flow in a network within a time interval; generating a sequence of likelihoods corresponding to sample values in the signal and based at least in part on a historical probability distribution of previously received sample values corresponding to amounts of data flow in the network, wherein a likelihood is a probability of occurrence of a corresponding sample value in the signal; identifying likelihood change points in the likelihood sequence by; selecting a parameter L corresponding to a minimum number of samples in a segment; appending L consecutive likelihoods to a buffer; computing a sequence of first sum values of the likelihoods in the buffer; obtaining a sequence of second sum values; determining the presence of a change point in the buffer based at least in part on a comparison between the first and second sum values, wherein a plurality of likelihoods preceding the change point have a first statistic value and a plurality of likelihoods following the change point have a second statistic value different from the first statistic value; and identifying a likelihood in the buffer as a change point based at least in part on the comparison; segmenting the discrete signal into a plurality of segments at samples corresponding to the identified change points such that a respective one of the samples corresponding to the identified change points is at one of a beginning or an end of each of the plurality of segments; identifying a segment as an anomaly based on a comparison between a statistic of the segment and a statistic of the historical probability distribution; and reconfiguring, responsive to identifying the segment as the anomaly, a component of the network. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9)
-
-
10. An apparatus for identifying an anomaly in a signal, comprising a processor and a memory unit storing computer executable instructions that when executed by the processor cause the processor to:
-
receive a discrete signal, having sample values corresponding to amounts of data flow in a network within a time interval; generate a sequence of likelihoods corresponding to sample values in the signal and based at least in part on a historical probability distribution of previously received sample values corresponding to amounts of data flow in the network, wherein a likelihood is a probability of occurrence of a corresponding sample value in the signal; identify likelihood change points in the likelihood sequence, by; selecting a parameter L corresponding to a minimum number of samples in a segment; appending L consecutive likelihoods to a buffer; computing a sequence of first sum values of the likelihoods in the buffer; obtaining a sequence of second sum values; determining the presence of a change point in the buffer based at least in part on a comparison between the first and second sum values, wherein a plurality of likelihoods preceding the change point have a first statistic value and a plurality of likelihoods following the change point have a second statistic value different from the first statistic value; and identifying a likelihood in the buffer as a change point based at least in part on the comparison; segment the discrete signal into a plurality of segments at samples corresponding to the identified change points such that a respective one of the samples corresponding to the identified change points is at one of a beginning or an end of each of the plurality of segments; identify a segment as an anomaly based on a comparison between a statistic of the segment and a statistic of the historical probability distribution; and reconfigure, responsive to identifying the segment as the anomaly, a component of the network. - View Dependent Claims (11, 12, 13, 14, 15, 16, 17, 18)
-
Specification