Properties of First-Level-Classifiers


FLC are derived from visualizations. Some inherited properties of them can be changed in the 'Transform' tab (see Properties of Visualizations). All other properties can be changed in the 'Properties' tab.





Property Value
Local Maxima Filter for data points that form a FLC-signature: (1) All points = All data points are copied to the signature. (2) x-maxima = Only maxima in rows are copied to the signature. (3) y-maxima = Only maxima in columns are copied to the signature. (3) xy-maxima = Only maxima of both rows and columns are copied to the signature.
x-Density Density filter for data points in rows of the signature.
y-Density Density filter for data points in columns of the signature.
Minimum Minimum threshold for elements of the positive (green) part of the signature.
Sensitivity Sensitivity of the algorithm that computes the correlation or distance function for the positive (green) part of a signature.
Max. x-Distance Maximum horizontal distance of elements of the negative (red) part of the signature from elements of the positive (green) part.
Max. y-Distance Maximum vertical distance of elements of the negative (red) part of the signature from elements of the positive (green) part.
Maximum Maximum threshold for elements of the negative (red) part of the signature.
Sensitivity Sensitivity of the algorithm that computes the correlation or distance function for the negative (red) part of a signature.
Correlation Basic correlation measure: (1) Euclidean Distance, (2) Cross-Correlation
Smooth Smooth the output of the combined correlation function (blue in the the FLC-Editor).
Compute Compute: (1) Positive Signature, (2) Positive Signature + Noise Floor
Threshold Minimum threshold for the combined correlation function (blue in the FLC-Editor) to generate an annotation.