In recent years, combination of multiple experts has become a major area of interest in designing practical and robust handwritten character recognition systems. Instead of building a single sophisticated and complicated classifier which is capable to handling all the various types of variations that are present in the handwritten character set, it has proved more prudent to apply relatively simpler classifiers (experts) by formulating ways of combining their individual decisions in order to generate robust and confident decisions. This paper presents a new class of decision combination approaches and compares the effectiveness of these approaches in successfully combining decisions by multiple experts in the specific application of handwritten numeral recognition. Although the proposed approaches have been applied to a specific task of handwritten numeral recognition, the underlying concepts are completely generalised and should be applicable to a very broad task domain.