Minimizing group generation in computer systems with limited computing resources

A method comprising:

• receiving a data set representing information regarding a plurality of attributes, wherein each of the plurality of attributes represents a corresponding plurality of measurements for a corresponding plurality of data items;
• determining, for each of the plurality of attributes, a plurality of overlapping matrices, representing overlapping ones of the corresponding plurality of measurements for each of the plurality of attributes;
• generating a plurality of samples of weights, wherein each of the plurality of samples of weights includes a corresponding weight for each of the plurality of overlapping matrices;
• determining a plurality of weighted overlap matrices by combining, for each of the plurality of samples of weights, a corresponding overlap matrix with the corresponding weight;
• determining a plurality of similarity matrices by combining, for each of the plurality of samples of weights, the plurality of weighted overlap matrices;
• performing a clustering analysis on the plurality of similarity matrices to identify, for each of the plurality of samples of weights, a corresponding plurality of desired groupings, wherein each of the corresponding plurality of desired groupings represents a corresponding group of the plurality of data items, and wherein the plurality of desired groupings across the plurality of similarity matrices comprise a set of groupings;
• and
• selecting a group from the set of groupings to obtain a selected grouping.

A system comprising:

• a processor having a defined amount of processing resources;
• and
• a data repository in communication with the processor, wherein the data repository stores:
• a plurality of attributes, each of the plurality of attributes representing a corresponding plurality of measurements for a corresponding plurality of data items;
• a plurality of overlapping matrices representing overlapping ones of the corresponding plurality of measurements for each of the plurality of attributes;
• a plurality of samples of weights, wherein each of the plurality of samples of weights includes a corresponding weight for each of the plurality of overlapping matrices;
• a plurality of weighted overlap matrices;
• a plurality of similarity matrices;
• corresponding pluralities of groupings representing groupings of the plurality of data items taken from a corresponding similarity matrix in the plurality of similarity matrices;
• a set of groupings representing the plurality of groupings across the plurality of similarity matrices;
• a selected grouping selected from the set of groupings;
• and
• an overlapping matrix generator configured to determine the plurality of overlapping matrices;
• a weight generator configured to generate the plurality of samples of weights;
• a similarity matrix generator configured to:
• determine the plurality of weighted overlap matrices by combining, for each of the plurality of samples of weights, a corresponding overlap matrix with the corresponding weight;
• and
• determine the plurality of similarity matrices by combining, for each of the plurality of samples of weights, the plurality of weighted overlap matrices;
• and
• a group selector configured to:
• perform a cluster analysis on the plurality of similarity matrices to identify, for each of the plurality of samples of weights, the corresponding pluralities of groupings;
• select the selected grouping from the set of groupings;
• and
• report the selected grouping.