Invention Grant
- Patent Title: Visual modification and training of an anomaly detection image
-
Application No.: US15087338Application Date: 2016-03-31
-
Publication No.: US10222959B2Publication Date: 2019-03-05
- Inventor: Serguei Mankovskii , Ian Davis , Michael Godfrey
- Applicant: Serguei Mankovskii , Ian Davis , Michael Godfrey
- Applicant Address: US NY New York
- Assignee: CA, Inc.
- Current Assignee: CA, Inc.
- Current Assignee Address: US NY New York
- Agency: Baker Botts L.L.P.
- Main IPC: G06F3/0484
- IPC: G06F3/0484 ; G06T11/20

Abstract:
A method includes formatting for display, on a visual screen, an image comprising: (1) a coordinate system, (2) a plurality of distinguishable areas within the coordinate system, each distinguishable area graphically representing a respective formula, and (3) a plurality of data points. The method also includes receiving user input comprising a modification to a particular distinguishable area. In response to receiving the user input, the method includes modifying one or more respective formulas based on the modification to the distinguishable area. For each data point, the method further includes associating the data point with one of the distinguishable areas by determining which of the modified formulas the data point falls within. The method further includes formatting for display a graphical representation of each modified formula. The method additionally includes storing the graphical representation of each modified formula for use as a modified image in future operations.
Public/Granted literature
- US20170285917A1 VISUAL MODIFICATION AND TRAINING OF AN ANOMALY DETECTION IMAGE Public/Granted day:2017-10-05
Information query