Weighted Statistics

Introduction

There are several different reasons why a statistical analysis needs to adjust for weighting. In literature reasons are mainly diveded in to groups.

The first group is when some of the measurements are known to be more precise than others. The more precise a measurement is, the larger weight it is given. The simplest case is when the weight are given before the measurements and they can be treated as deterministic. It becomes more complicated when the weight can be determined not until afterwards, and even more complicated if the weight depends on the value of the observable.

The second group of situations is when calculating averages over one distribution and sampling from another distribution. Compensating for this discrepency weights are introduced to the analysis. A simple example may be that we are interviewing people but for economical reasons we choose to interview more people from the city than from the countryside. When summarizing the statistics the answers from the city are given a smaller weight. In this example we are choosing the proportions of people from countryside and people from city being intervied. Hence, we can determine the weights before and consider them to be deterministic. In other situations the proportions are not deterministic, but rather a result from the sampling and the weights must be treated as stochastic and only in rare situations the weights can be treated as independent of the observable.

Since there are various origins for a weight occuring in a statistical analysis, there are various ways to treat the weights and in general the analysis should be tailored to treat the weights correctly. We have not chosen one situation for our implementations, so see specific function documentation for what assumtions are made. Though, common for implementations are the following:

The last point implies that a data point with zero weight is ignored also when the value is NaN. An important case is when weights are binary (either 1 or 0). Then we get the same result using the weighted version as using the data with weight not equal to zero and the non-weighted version. Hence, using binary weights and the weighted version missing values can be treated in a proper way.

AveragerWeighted

Mean

For any situation the weight is always designed so the weighted mean is calculated as $ m=\frac{\sum w_ix_i}{\sum w_i} $, which obviously fulfills the conditions above.

In the case of varying measurement error, it could be motivated that the weight shall be $ w_i = 1/\sigma_i^2 $. We assume measurement error to be Gaussian and the likelihood to get our measurements is $ L(m)=\prod (2\pi\sigma_i^2)^{-1/2}e^{-\frac{(x_i-m)^2}{2\sigma_i^2}} $. We maximize the likelihood by taking the derivity with respect to $ m $ on the logarithm of the likelihood $ \frac{d\ln L(m)}{dm}=\sum \frac{x_i-m}{\sigma_i^2} $. Hence, the Maximum Likelihood method yields the estimator $ m=\frac{\sum w_i/\sigma_i^2}{\sum 1/\sigma_i^2} $.

Variance

In case of varying variance, there is no point estimating a variance since it is different for each data point.

Instead we look at the case when we want to estimate the variance over $f$ but are sampling from $ f' $. For the mean of an observable $ O $ we have $ \widehat O=\sum\frac{f}{f'}O_i=\frac{\sum w_iO_i}{\sum w_i} $. Hence, an estimator of the variance of $ X $ is

$ s^2 = <X^2>-<X>^2= $

$ = \frac{\sum w_ix_i^2}{\sum w_i}-\frac{(\sum w_ix_i)^2}{(\sum w_i)^2}= $

$ = \frac{\sum w_i(x_i^2-m^2)}{\sum w_i}= $

$ = \frac{\sum w_i(x_i^2-2mx_i+m^2)}{\sum w_i}= $

$ = \frac{\sum w_i(x_i-m)^2}{\sum w_i} $

This estimator fulfills that it is invariant under a rescaling and having a weight equal to zero is equivalent to removing the data point. Having all weights equal to unity we get $ \sigma=\frac{\sum (x_i-m)^2}{N} $, which is the same as returned from Averager. Hence, this estimator is slightly biased, but still very efficient.

Standard Error

The standard error squared is equal to the expexted squared error of the estimation of $m$. The squared error consists of two parts, the variance of the estimator and the squared bias:

$ <m-\mu>^2=<m-<m>+<m>-\mu>^2= $ $ <m-<m>>^2+(<m>-\mu)^2 $.

In the case when weights are included in analysis due to varying measurement errors and the weights can be treated as deterministic, we have

$ Var(m)=\frac{\sum w_i^2\sigma_i^2}{\left(\sum w_i\right)^2}= $ $ \frac{\sum w_i^2\frac{\sigma_0^2}{w_i}}{\left(\sum w_i\right)^2}= $ $ \frac{\sigma_0^2}{\sum w_i}, $

where we need to estimate $ \sigma_0^2 $. Again we have the likelihood

$ L(\sigma_0^2)=\prod\frac{1}{\sqrt{2\pi\sigma_0^2/w_i}}\exp{(-\frac{w_i(x-m)^2}{2\sigma_0^2})} $ and taking the derivity with respect to $\sigma_o^2$,

$ \frac{d\ln L}{d\sigma_i^2}= $ $ \sum -\frac{1}{2\sigma_0^2}+\frac{w_i(x-m)^2}{2\sigma_0^2\sigma_o^2} $

which yields an estimator $ \sigma_0^2=\frac{1}{N}\sum w_i(x-m)^2 $. This estimator is not ignoring weights equal to zero, because deviation is most often smaller than the expected infinity. Therefore, we modify the expression as follows $\sigma_0^2=\frac{\sum w_i^2}{\left(\sum w_i\right)^2}\sum w_i(x-m)^2$ and we get the following estimator of the variance of the mean $\sigma_0^2=\frac{\sum w_i^2}{\left(\sum w_i\right)^3}\sum w_i(x-m)^2$. This estimator fulfills the conditions above: adding a weight zero does not change it: rescaling the weights does not change it, and setting all weights to unity yields the same expression as in the non-weighted case.

In a case when it is not a good approximation to treat the weights as deterministic, there are two ways to get a better estimation. The first one is to linearize the expression $\left<\frac{\sum w_ix_i}{\sum w_i}\right>$. The second method when the situation is more complicated is to estimate the standard error using a bootstrapping method.

AveragerPairWeighted

Here data points come in pairs (x,y). We are sampling from $f'_{XY}$ but want to measure from $f_{XY}$. To compensate for this decrepency, averages of $g(x,y)$ are taken as $\sum \frac{f}{f'}g(x,y)$. Even though, $X$ and $Y$ are not independent $(f_{XY}\neq f_Xf_Y)$ we assume that we can factorize the ratio and get $\frac{\sum w_xw_yg(x,y)}{\sum w_xw_y}$

Covariance

Following the variance calculations for AveragerWeighted we have $Cov=\frac{\sum w_xw_y(x-m_x)(y-m_y)}{\sum w_xw_y}$ where $m_x=\frac{\sum w_xw_yx}{\sum w_xw_y}$

Correlation

As the mean is estimated as $ m_x=\frac{\sum w_xw_yx}{\sum w_xw_y} $, the variance is estimated as $ \sigma_x^2=\frac{\sum w_xw_y(x-m_x)^2}{\sum w_xw_y} $. As in the non-weighted case we define the correlation to be the ratio between the covariance and geometrical average of the variances

$ \frac{\sum w_xw_y(x-m_x)(y-m_y)}{\sqrt{\sum w_xw_y(x-m_x)^2\sum w_xw_y(y-m_y)^2}} $.

This expression fulfills the following

Score

Pearson

$\frac{\sum w(x-m_x)(y-m_y)}{\sqrt{\sum w(x-m_x)^2\sum w(y-m_y)^2}}$.

See AveragerPairWeighted correlation.

ROC

An interpretation of the ROC curve area is the probability that if we take one sample from class $+$ and one sample from class $-$, what is the probability that the sample from class $+$ has greater value. The ROC curve area calculates the ratio of pairs fulfilling this

$ \frac{\sum_{\{i,j\}:x^-_i<x^+_j}1}{\sum_{i,j}1}. $

An geometrical interpretation is to have a number of squares where each square correspond to a pair of samples. The ROC curve follows the border between pairs in which the samples from class $+$ has a greater value and pairs in which this is not fulfilled. The ROC curve area is the area of those latter squares and a natural extension is to weight each pair with its two weights and consequently the weighted ROC curve area becomes

$ \frac{\sum_{\{i,j\}:x^-_i<x^+_j}w^-_iw^+_j}{\sum_{i,j}w^-_iw^+_j} $

This expression is invariant under a rescaling of weight. Adding a data value with weight zero adds nothing to the exprssion, and having all weight equal to unity yields the non-weighted ROC curve area.

tScore

Assume that $x$ and $y$ originate from the same distribution $N(\mu,\sigma_i^2)$ where $\sigma_i^2=\frac{\sigma_0^2}{w_i}$. We then estimate $\sigma_0^2$ as $ \frac{\sum w(x-m_x)^2+\sum w(y-m_y)^2} {\frac{\left(\sum w_x\right)^2}{\sum w_x^2}+ \frac{\left(\sum w_y\right)^2}{\sum w_y^2}-2} $ The variance of difference of the means becomes $ Var(m_x)+Var(m_y)=\\\frac{\sum w_i^2Var(x_i)}{\left(\sum w_i\right)^2}+\frac{\sum w_i^2Var(y_i)}{\left(\sum w_i\right)^2}= \frac{\sigma_0^2}{\sum w_i}+\frac{\sigma_0^2}{\sum w_i}, $ and consequently the t-score becomes $ \frac{\sum w(x-m_x)^2+\sum w(y-m_y)^2} {\frac{\left(\sum w_x\right)^2}{\sum w_x^2}+ \frac{\left(\sum w_y\right)^2}{\sum w_y^2}-2} \left(\frac{1}{\sum w_i}+\frac{1}{\sum w_i}\right), $

For a $w_i=w$ we this expression get condensed down to $ \frac{w\sum (x-m_x)^2+w\sum (y-m_y)^2} {n_x+n_y-2} \left(\frac{1}{wn_x}+\frac{1}{wn_y}\right), $ in other words the good old expression as for non-weighted.

FoldChange

Fold-Change is simply the difference between the weighted mean of the two groups $\frac{\sum w_xx}{\sum w_x}-\frac{\sum w_yy}{\sum w_y}$

WilcoxonFoldChange

Taking all pair samples (one from class $+$ and one from class $-$) and calculating the weighted median of the distances.

Distance

A Distance measures how far apart two ranges are. A Distance should preferably meet some criteria:

Weighted Distance

Weighted Distance is an extension of usual unweighted distances, in which each data point is accompanied with a weight. A weighted distance should meet some criteria:

The last condition, duplicate property, implies that setting a weight to zero is not equivalent to removing the data point. This behavior is sensible because otherwise we would have a bias towards having ranges with small weights being close to other ranges. For a weighted distance, meeting these criteria, it might be difficult to show that the triangle inequality is fulfilled. For most algorithms the triangle inequality is not essential for the distance to work properly, so if you need to choose between fulfilling triangle inequality and these latter criteria it is preferable to meet the latter criteria.

In test/distance_test.cc there are tests for testing these properties.

Kernel

Polynomial Kernel

The polynomial kernel of degree $N$ is defined as $(1+<x,y>)^N$, where $<x,y>$ is the linear kernel (usual scalar product). For the weighted case we define the linear kernel to be $<x,y>=\frac{\sum {w_xw_yxy}}{\sum{w_xw_y}}$ and the polynomial kernel can be calculated as before $(1+<x,y>)^N$.

Gaussian Kernel

We define the weighted Gaussian kernel as $\exp\left(-N\frac{\sum w_xw_y(x-y)^2}{\sum w_xw_y}\right)$.

Regression

Naive

Linear

We have the model

$ y_i=\alpha+\beta (x-m_x)+\epsilon_i, $

where $\epsilon_i$ is the noise. The variance of the noise is inversely proportional to the weight, $Var(\epsilon_i)=\frac{\sigma^2}{w_i}$. In order to determine the model parameters, we minimimize the sum of quadratic errors.

$ Q_0 = \sum \epsilon_i^2 $

Taking the derivity with respect to $\alpha$ and $\beta$ yields two conditions

$ \frac{\partial Q_0}{\partial \alpha} = -2 \sum w_i(y_i - \alpha - \beta (x_i-m_x)=0 $

and

$ \frac{\partial Q_0}{\partial \beta} = -2 \sum w_i(x_i-m_x)(y_i-\alpha-\beta(x_i-m_x)=0 $

or equivalently

$ \alpha = \frac{\sum w_iy_i}{\sum w_i}=m_y $

and

$ \beta=\frac{\sum w_i(x_i-m_x)(y-m_y)}{\sum w_i(x_i-m_x)^2}=\frac{Cov(x,y)}{Var(x)} $

Note, by having all weights equal we get back the unweighted case. Furthermore, we calculate the variance of the estimators of $\alpha$ and $\beta$.

$ \textrm{Var}(\alpha )=\frac{w_i^2\frac{\sigma^2}{w_i}}{(\sum w_i)^2}= \frac{\sigma^2}{\sum w_i} $

and $ \textrm{Var}(\beta )= \frac{w_i^2(x_i-m_x)^2\frac{\sigma^2}{w_i}} {(\sum w_i(x_i-m_x)^2)^2}= \frac{\sigma^2}{\sum w_i(x_i-m_x)^2} $

Finally, we estimate the level of noise, $\sigma^2$. Inspired by the unweighted estimation

$ s^2=\frac{\sum (y_i-\alpha-\beta (x_i-m_x))^2}{n-2} $

we suggest the following estimator

$ s^2=\frac{\sum w_i(y_i-\alpha-\beta (x_i-m_x))^2}{\sum w_i-2\frac{\sum w_i^2}{\sum w_i}} $


Generated on Tue Jan 18 02:20:10 2011 for yat by  doxygen 1.5.5