Democracy in pattern classifications: combinations of votes from various pattern classifiers

The objective of this paper is to show that a combination of votes from various pattern classifiers is better than a single vote from each individual classifier. A proposed support function is used in the combination of votes. The combination of outputs is motivated by the fact that decisions made by teams are generally better than those made by individuals. The decision maker at the outputs of the front-end classifiers is called the combined classifier (CC). The proof of the theory of the combining method is obtained using the principle of mathematical induction. Experimental investigation has been conducted to verify the theory. The first experiment was conducted using 5000 training digits and the second experiment was conducted using 10000 training digits. CC achieved a recognition accuracy of 86.67% compared with 70% of the best individual classifier in the second experiment. The results show that the theoretical and the experimental values are in good agreement.