US 12,437,029 B1
System and method for identifying outlier data and generating corrective action
Blake Browder, Dallas, TX (US); and Joy Figarsky, Little Rock, AR (US)
Assigned to Signet Health Corporation, North Richland Hills, TX (US)
Filed by Signet Health Corporation, North Richland Hills, TX (US)
Filed on Nov. 24, 2024, as Appl. No. 18/957,803.
Int. Cl. G06F 16/35 (2025.01); G06F 18/232 (2023.01); G06F 21/62 (2013.01)
CPC G06F 18/232 (2023.01) [G06F 21/6218 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A system for identifying outlier data and generating corrective action, wherein the system comprises:
at least a processor; and
a memory communicatively connected to the at least a processor, wherein the memory contains instructions configuring the at least a processor to:
obtain a dataset comprising a plurality of data points;
apply a clustering model to the dataset to generate a set of clusters based on inherent relationships between the data points;
determine a distance metric for each data point relative to its corresponding cluster centroid;
define an outlier threshold based on the distance metrics of the data points;
identify the data points that exceed the outlier threshold as one or more outliers;
classify the one or more outliers across one or more axes, wherein classifying the one or more outliers further comprises categorizing each of the one or more outliers into specific outlier types;
correlate each categorized specific outlier type to a recommended corrective action using a look-up table; and
output a report of the one or more identified outliers and the specific outlier types, wherein the report suggests the recommended corrective actions and insights for each of the one or more identified outliers.