A learning contour identification system using a portable contour metric derived from contour mapping.

摘要

By converting the data format to a contour metric and then converting each contour line of the mapping to a set of contour pattern metrics, each metric created is an iteration of the learning-type contour line identification system defined herein. Systems and methods are provided having a level of contour representation that is presented for each. This conversion of data instances to contour metrics is determined by the user for other types and functions of machines by the learning contour identification system for further analysis of the patterns discovered and labeled by the system. So that the relevant data of the dataset can be obtained. The present invention is not limited to being performed by a data format representation for signal, image, or waveform embodiments to identify and track patterns of amplitude, frequency, phase, and density functions within a data case, Or by detecting and then using a combination of statistics, feedback adaptation, classification, training algorithm metrics stored in hardware to identify patterns in past data cases that are repeated in the future, or in current data cases, A high percentage of labeling and identification is thereby achieved. [Selection] Figure 1