FIND ME ON

GitHub

LinkedIn

Centroid Condition for MSE

🌱

Theorem
InfoTheory

The yiy_{i}’s minimizing the MSE distortion given a partition {R1,,RN}\{ R_{1},\dots,R_{N} \} are uniquely given by yi=E[XXRi], i=1,,Ny_{i}=E[X|X\in R_{i}], \ i=1,\dots,N

The yiy_{i}’s minimizing the MSE Vector Quantizer given a partition {R1,,RN}\{ R_{1},\dots,R_{N} \} are uniquely given by ci=E[XXRi], i=1,,N\mathbf{c}_{i}=E[\mathbf{X}|\mathbf{X}\in R_{i}], \ i=1,\dots,N

If we have that XfX\sim f, then fXRi(x)={f(x)P(XRi)if xRi0otherwisef_{X|R_{i}}(x)=\begin{cases} \frac{f(x)}{P(X\in R_{i})} & \text{if }x\in R_{i} \\ 0 & \text{otherwise} \end{cases}so E[XXRi]=xfXRi(x)dx=Rixf(x)dxRif(x)dx=Rixf(x)dxP(XRi)E[X|X\in R_{i}]=\int\limits _{-\infty}^{\infty}xf_{X|R_{i}}(x) \, dx =\frac{ \int\limits _{R_{i}}xf(x) \, dx }{\int\limits_{R_{i}} f(x) \, dx }=\frac{ \int\limits _{R_{i}}xf(x) \, dx }{P(X\in R_{i})}