摘要：Mathematical programming is a backbone of operations research and as such has led to the solution of multiple optimization problems in a variety of fields. When this research concerns pure or applied science, it is sometimes refered to as discovery science. Some results obtained in the last decades at GERAD will be presented and discussed. We will answer the following questions:

(i) Did George Boole invent probabilistic logic? (ii) Did he invent linear programming? (iii) Does the stability of fullerenes increase with the number of Kékulé structures? (iv) Is Vincke's wife's octagone suboptimal? (v) Can the AutoGraphiX system, developped at GERAD Montreal, generate over 1000 conjectures in graph theory? (vi) Can Chemical graphs with maximum or minimum Randic or other topological indices be determined by linear 0-1-programming?