1: \begin{abstract}

2:        We consider the rate of convergence of the expected loss of empirically optimal vector quantizers. Earlier results show that the mean-squared expected distortion for any fixed distribution supported on a bounded set and satisfying some regularity conditions decreases at the rate $\mathcal{O}(\log n/n)$. We prove that this rate is actually $\mathcal{O}(1/n)$. Although these conditions are hard to check, we show that well-polarized distributions with continuous densities supported on a bounded set are included in the scope of this result. 

3:          

4: \end{abstract}