Issue No. 06 - June (2003 vol. 25)
Frank J. Iannarilli Jr. , IEEE
<p><b>Abstract</b>—We reformulate branch-and-bound feature selection employing <em>L_\infty</em> or particular <em>L_p</em> metrics, as mixed-integer linear programming (MILP) problems, affording convenience of widely available MILP solvers. These formulations offer direct influence over individual pairwise interclass margins, which is useful for feature selection in multiclass settings.</p>
Feature selection, discrimination, classification, mixed-integer linear programming, branch-and-bound.
P. A. Rubin and F. J. Iannarilli Jr., "Feature Selection for Multiclass Discrimination via Mixed-Integer Linear Programming," in IEEE Transactions on Pattern Analysis & Machine Intelligence, vol. 25, no. , pp. 779-783, 2003.