Evaluating anomaly for one-class classifiers in machine condition monitoring
View Patent ↗A method for monitoring machine conditions provides additional information using a one-class classifier in which an evaluation function is learned. In the method, a distance is determined from an anomaly measurement x to a boundary of a region R 1 containing all acceptable measurements. The distance is used as a measure of the extent of the anomaly. The distance is found by searching along a line from the anomaly to a closest acceptable measurement within the region R 1 .
1. A method for evaluating an anomaly measurement x′ in a machine condition monitoring system including a processor, wherein measurements x i are evaluated in a one-class classifier having a decision region R 1 for the class C 1 such that an evaluation function ƒ(x) is greater than or equal to a threshold T for a measurement x within the region, and less than the threshold T outside the region, the method comprising the steps of:
training the one-class classifier to establish the decision region R 1 from a set of training samples {x 1 , x 2 , . . . , x N };
receiving the anomaly measurement x′;
determining that the anomaly measurement is outside the region R 1 ;
determining, using the processor, a distance from the measurement x′ to a boundary of the region R 1 ; and
evaluating the anomaly measurement x′ based on the distance;
wherein the step of determining a distance from the measurement x′ to the region R 1 is performed iteratively.
2. The method of claim 1 , wherein the distance is a Euclidean distance.
3. The method of claim 1 , further comprising the step of:
obtaining the evaluation function ƒ(x) using an algorithm selected from the group consisting of a nearest neighbor algorithm, a probability density function and an inner product for a support vector representation machine.
4. A non-transitory computer-usable medium having computer readable instructions stored thereon for execution by a processor to perform a method evaluating an anomaly measurement x′ in a machine condition monitoring system wherein measurements x i are evaluated in a one-class classifier having a decision region R 1 for the class C 1 such that an evaluation function ƒ(x) is greater than or equal to a threshold T for a measurement x within the region, and less than the threshold T outside the region, the method comprising the steps of:
training the one-class classifier to establish the decision region R 1 from a set of training samples {x 1 , x 2 , . . . , x N };
receiving the anomaly measurement x′;
determining that the anomaly measurement is outside the region R 1 ;
determining a distance from the measurement x′ to a boundary of the region R 1 ; and
evaluating the anomaly measurement x′ based on the distance;
wherein the step of determining a distance from the measurement x′ to the region R 1 is performed iteratively.
5. The non-transitory computer-usable medium of claim 4 , wherein the distance is a Euclidean distance.
6. The non-transitory computer-usable medium of claim 4 , further comprising the step of:
obtaining the evaluation function ƒ(x) using an algorithm selected from the group consisting of a nearest neighbor algorithm, a probability density function and an inner product for a support vector representation machine.