Multiclass neural network classifiers are typically trained using cross-entropy loss. Following training, the performance of this same neural network is evaluated using an application-specific metric based on the multiclass confusion matrix, such as the Macro -Score. It is questionable whether the use of cross-entropy will yield a classifier that aligns with the intended application-specific performance criteria, particularly in scenarios where there is a need to emphasize one aspect of classifier performance. For example, if greater precision is preferred over recall, the value in the evaluation metric can be adjusted accordingly, but the cross-entropy objective remains unaware of this preference during training. We propose a method that addresses this training-evaluation gap for multiclass neural network classifiers such that users can train these models informed by the desired final -Score. Following prior work in binary classification, we utilize the concepts of the soft-set confusion matrices and a piecewise-linear approximation of the Heaviside step function. Our method extends the binary soft-set confusion matrix to a multiclass confusion matrix and proposes dynamic adaptation of the threshold value , which parameterizes the piecewise-linear Heaviside approximation during run-time. We present a theoretical analysis that shows that our method can be used to optimize for a soft-set based approximation of Macro- that is a consistent estimator of Macro-, and our extensive experiments show the practical effectiveness of our approach.