THE BOND-ENERGY ALGORITHM REVISITED

The bond energy algorithm of McCormick, Schweitzer, and White [1] is examined in the context of related strategies of data analysis that seek to solve problems in production research, imaging, and related engineering problems. A taxonomy of types of input data and forms of matrix structure, adopted from other areas of data analysis, serves to clarify some distinctions that have been at most implicit in published alternatives to the bond energy approach. We consider the objective function initially proposed for this approach and deduce some of its properties and resulting limitations. We offer some extensions of the original technique and criticize one alternative proposed by King [2]. © 1990 IEEE