Performance analysis of pattern classifier combination by plurality voting

Plurality voting is widely used in pattern recognition practice. However, there is little theoretical analysis of plurality voting. In this paper, we attempt to explore the rationales behind plurality voting. The recognition/error/rejection rates of plurality voting are compared with those of majority voting under different conditions. It is demonstrated that plurality voting is more efficient in achieving the tradeoff between rejection rate and error rate. We also discuss some practical problems when applying plurality voting to real-world applications.