Improving kNN classifier performance using Quantized Cell Method (QCM)

Nowadays improving classification accuracy and speed of test are two importantfields of researchers’ studies, and even a little increase in these classifier characteristics isvaluable. In this article a kNN-based method is introduced which is called quantized cellmethod” (QCM). This method contains two phases: (1) forward phase, (2) backward phase.These two phases includes quantization and encoding steps. QCM-kNN modifies traditionalkNN algorithm by adding a train phase, and experimental results exhibits that thismodification can improve both the classification accuracy and test speed.

Keywords:

k-Nearest-Neighbor , kNN search algorithm , Classification , Quantization. Artificial intelligence