Questions tagged [mathematical-statistics]
Mathematical theory of statistics, concerned with formal definitions and general results.
2,269
questions with no upvoted or accepted answers
20
votes
0
answers
882
views
Is there a general expression for ancillary statistics in exponential families?
An i.i.d sample $X_1,\dots,X_n$ from a scale family with c.d.f. $F(\frac{x}{\sigma})$ has $S(X)$ as an ancillary statistic if $S(X)$ depends on the sample only through $\frac{X_1}{X_n},\cdots,\frac{X_{...
11
votes
0
answers
2k
views
What does it mean to take the expectation with respect to a probability distribution?
I see this expectation in a lot of machine learning literature:
$$\mathbb{E}_{p(\mathbf{x};\mathbf{\theta})}[f(\mathbf{x};\mathbf{\phi})] = \int p(\mathbf{x};\mathbf{\theta}) f(\mathbf{x};\mathbf{\phi}...
9
votes
1
answer
12k
views
What does "def" above an equals sign mean?
I am reading this:
https://www.cs.ubc.ca/~murphyk/Papers/bayesGauss.pdf
and on equation (17), there is a def on top of the equal sign. What does this mean?
9
votes
0
answers
2k
views
Taylor Series and Multivariate Delta Method
I asked this question on https://math.stackexchange.com/ but did not get any answer. Sorry for cross posting.
I'm trying to understand delta method for matrices and vectors to find the variance-...
8
votes
0
answers
777
views
Intuitive explanation for Marchenko-Pastur law
I am looking for an intuitive reasoning behind the Marchenko Pastur law, which is described as a law of large numbers analog for random matrices. I know the law gives the probability density function ...
8
votes
0
answers
1k
views
Proof of Kolmogorov-Smirnov test
Could someone provide me a reference, preferably a book, where I can find detailed proofs and explanations of the Kolmogorov-Smirnov test (including the two-sample variant) and the derivation of the $...
8
votes
0
answers
142
views
Orthogonal intersection in a Riemannian manifold
Let $S$ be the set of all probability distributions on $\mathbb{R}$ and $S_n=\{p_\theta\}$ be an $n$ dimensional submanifold of parameterized family of probability distributions on $\mathbb{R}$ where $...
8
votes
1
answer
242
views
Finding the distribution of sample range for a Beta population
Let $X_1,X_2,\ldots,X_n$ be i.i.d random variables having density
$$f(x)=2(1-x)\mathbf1_{0<x<1}$$
I am trying to derive the distribution of the sample range $R=X_{(n)}-X_{(1)}$.
The usual way I ...
7
votes
0
answers
367
views
Sklar’s Extension Theorem and support restrictions
This question is about an application of the Sklar's Extension Theorem, whose proof can be found in Sklar, A. (1996), "Random variables, distribution functions, and copulas: A personal look ...
7
votes
0
answers
109
views
Proof: Nearest Neighbor classifier achieves Bayes rate asymptotically on countable domains
I am trying to understand in which situations the 1-NN classifier asymptotically attains the Bayes error rate. My intuition is that if the domain is countable, then 1-NN will asymptotically do as well ...
7
votes
0
answers
51
views
Simulate correlate random variables with given marginal distribution where one is always larger
Is it possible to simulate pairs of random variables with a given marginal distribution and population correlation where one random variable is larger than the other?
More formally, I need to simulate ...
7
votes
0
answers
128
views
How do we call a more extreme case of fat tails than a power law?
According to Wikipedia the most extreme case of a fat tail follows a power law:
The most extreme case of a fat tail is given by a distribution whose tail decays like a power law.
That is, if the ...
7
votes
0
answers
105
views
Inequalities on Fisher Information / expected second derivative?
Under some regularity conditions we can compute fisher information as
$ - \mathbb{E}_{\theta_0} [\frac{\partial}{\partial \theta^2} \ln f(x;\theta_0)] $
I was wondering if there are some kind of ...
7
votes
0
answers
314
views
Strange connection between Bernouilli, Uniform and Geometric distributions
Final update on 11/29/2019: I have worked on this a bit more, and wrote an article summarizing all the main findings. You can read it here.
Let us consider $Z = X_1 + X_1 X_2 + X_1 X_2 X_3 +\cdots$ ...
7
votes
0
answers
97
views
French website Providing Instruction/Tutorials on Statistical Theory
This is somewhat of an odd question for CV, but since it's a question about statistical education, I think it falls within the scope of CV. Several years ago I stumbled across a French website that ...
7
votes
0
answers
224
views
Time evolution of a Bayesian posterior
I have a question regarding the time evolution of a quantity related to a Bayesian posterior.
Suppose we have binary parameter space $\{ s_1, s_2 \}$ with prior $(p, 1-p)$,
The data generating ...
7
votes
0
answers
234
views
Cox's Theorem: the necessity of (un)countably additivity
I've been trying to understand Cox's Theorem and the problems surrounding it. There's so much information on this topic that I've become confused as to the exact state of the theorem. I've gathered ...
7
votes
0
answers
234
views
Cox's Theorem: ignorance, objective priors, and the Mind Projection Fallacy
I've been trying to understand Cox's Theorem and the problems surrounding it. There's so much information on this topic that I've become confused as to the exact state of the theorem. I've gathered ...
7
votes
0
answers
3k
views
Help with a proof of Bayes classifier optimality
I have a class assignment to provide a proof that Bayes classifier for the two label version is optimal in that it's error rate is always ${\le}$ any other classifier.
I've never worked through a ...
7
votes
1
answer
390
views
minimizer weighted linear regression
In a regression problem, with $y=X\theta+\epsilon$ and $X$ is an $n$ by $p$ matrix
the ‘weighted least squares estimate is the minimizer $\theta^{*}$ of $f(\theta)=\sum_{i=1}^{n}\omega_{i}(y_i-x_i^{'}\...
7
votes
0
answers
3k
views
Deriving the maximum likelihood for a generative classification model for K classes
In Christopher Bishop's book "Pattern Recognition and Machine learning", there is the following question:
Consider a generative classification model for $K$ classes defined by the prior class ...
7
votes
0
answers
275
views
Reference Request: Information Geometry for Ridge Regression
I am reading the book "regression estimators" by Gruber 2010 where he uses this technique to compare Ridge Regressors, however he concentrates on deriving the mathematical results without ...
7
votes
0
answers
162
views
Hypothesis test on the Euclidean length of an unknown vector
Question
Suppose I observe a vector $\mathbf{x}=[X_1 \ldots X_n]$, where each $X_i=m_i+n_i$, with $n_i$ being an independent zero-mean Gaussian random variable with variance $\sigma^2$ (i.e. $n_i\sim\...
6
votes
0
answers
169
views
Distribution that doesn't belong to any maximum domain of attraction?
Question
Does there exist a (non-degenerate) distribution that does NOT belong to any maximum domain of attraction?
That is:
Does there exist any non-degenerate probability distribution function $F$ ...
6
votes
0
answers
80
views
Symbolic Formulae for Linear Mixed Models
I would like to understand how to create a good formula for a linear mixed model, using the Machines data set from the package ...
6
votes
0
answers
126
views
Does Fisher scoring always outperform Newton optimization?
My understanding is that Fisher scoring has several advantages over Newton raphson optimization such as
Computational efficiency: if certain conditions are met (example:During MLE estimation, if link ...
6
votes
0
answers
268
views
Why is $X$ not an identifiable statistical model
In my textbook, Identifiablity is defined as so:
For any $\theta_1, \theta_2 \in \Theta$ , if $\theta_1 \neq \theta_2 \Rightarrow \Bbb P_{\theta_1} \neq \Bbb P_{\theta_2}$ , where $\Bbb P_{\theta}$ ...
6
votes
2
answers
107
views
How to win this dice probability game?
The game is a variation of Pig. Here is how the game works:
There are about 20 players. Each round, a single six sided die is rolled. All players add that rolled number to their "bank." However, if a ...
6
votes
0
answers
517
views
Why we really need the concept of "Local" Rademacher complexity?
Recently, I have been studying High-Dimensional Statistics: A Non-Asymptotic Viewpoint written by Martin J. Wainwright. In this book, the author uses a special complexity measure which is called Local ...
5
votes
0
answers
59
views
Nuisance parameters and $o_p(n^{-1/4})$ convergence: citation
I'm looking for an original reference to a proof idea.
Suppose we have $n$ iid observations $(X_i,Y_i)$ and an estimating function
$$\bar U(\beta;\alpha)=\frac{1}{n}U(\beta;\alpha; X_i,Y_i)$$
where we ...
5
votes
0
answers
293
views
Is an inadmissible estimator necessarily dominated by some admissible estimator
Basic example: $X$ has a $p$-variate iid standard Normal distribution; the sample mean is not admissible if $p>2$ and is dominated by the Stein shrinkage estimator.
However, the Stein shrinkage ...
5
votes
1
answer
2k
views
How do I get the p value of AD test using the results of scipy.stats.anderson()
I am conducting an Anderson Darling test for normality using scipy.stats.anderson() command in python. I am getting test statistic, critical values at various ...
5
votes
0
answers
496
views
Why can't the complete class theorem be easily generalized to all locally-compact spaces?
So I was reading Christian P. Robert's The Bayesian Choice, going through the constellation of results related to complete class theorems, and I don't see why all of them are necessary. In particular, ...
5
votes
0
answers
265
views
Covariance of order statistics convergence?
Suppose I have a sample $(X_1 \dots X_n)$ and $(Y_1 \dots Y_n)$, all of which are $N(0,1)$ random variables. I am interested in the asymptotic behaviour of
$$\frac{1}{n} \sum_{i=0}^n X_{(i)}Y_{(i)} $$...
5
votes
0
answers
260
views
Combine absolute and relative difference into one metric
Imagine we're looking to buy some commodity, say a table, and we want to make a good deal. We know the "true" price of the tables at offer and we're interested in both the absolute price difference ...
5
votes
0
answers
2k
views
Calculating a confidence interval for a weighted sample
In a nutshell, I'd like to compute a confidence interval for some weighted sample day where the final value I'm seeking is a sum of different weighted samples. Illustrated here with a toy example set ...
5
votes
0
answers
176
views
Finding the Cauchy Principal Value Mean of the pdf for $Z=XY$ where $\mathbb{E}[Y]$ does not exist
Let's first define the Cauchy Principal Value Mean (PVM). For a continuous random variable $V$ with pdf $f_{V}(v)$, the PVM of $f_{V}$ is
\begin{equation}
\mathrm{PV}(\mathbb{E}[V])=\lim_{a\to\infty}\...
5
votes
0
answers
310
views
Variance of quotient of Poisson random variable and sum of the Poisson sample
Let
$$Y_1\sim \operatorname{Poisson}(\lambda_1)\\Y_2\sim \operatorname{Poisson}(\lambda_2),$$ where $Y_1$ and $Y_2$ are independent, and $\lambda_1, \lambda_2>0$.
What is the variance of $$\frac{...
5
votes
0
answers
863
views
Under what conditions will a Bayesian posterior fail to converge to a point mass?
Let's say you have a Bayesian model:
$$\theta' \sim g(\theta|\mu) $$
$$ y \sim p(y|\theta')$$
And we have some data ($n$ data points) $\mathbf{y}_n$, which we will use to perform inference on $\...
5
votes
0
answers
64
views
Why is the $\chi^{2}$ approximation for deviance GLM $\sim \operatorname{Binomial}(n_{i},\pi_{i})$ not valid when $n_{i} = 1$?
I know from McCullagh & Nelder's text (p.118) that the $\chi^{2}$ approximation for deviance for the binomial family is based on a limiting operation in which $n$, the number of observations, is ...
5
votes
0
answers
369
views
Calculating Formula with logs and Weighted Average
I am trying to figure out an equation from the following paper by Cadena and Kovak (2016): http://pubs.aeaweb.org/doi/pdfplus/10.1257/app.20140095 on pages 264 and 265. I don't think the context is ...
5
votes
0
answers
275
views
I'm not asking for a conjugate prior. Is there a distribution $p(x|y)$ that satisfies $\int p(x|y)Beta(y|a,b) dy = Beta(x| a', b')$?
I know the result of integrating a Gaussian against another Gaussian is still Gaussian, $$\int N(x|\mu_y,\sigma_y)N(y|\mu,\sigma) dy = N(x|\mu',\sigma')\quad.$$
Can I get the same form for Beta ...
5
votes
1
answer
413
views
Why low rank expansions can exploit the redundancy that exist between different feature channels and filters?
I read Jaderberg et al., 2014 paper about Speeding up Convolutional Neural Network with Low Rank Expansions. In the introduction, it is written in bold font:
Our key insight is to exploit the ...
5
votes
0
answers
1k
views
Conditional expectation everywhere non-zero while unconditional one zero?
I have a real-valued random variable $X$ that takes on positive and negative values, and
$$E(X)=0 \tag{A}$$
There is also another real-valued random variable $Y$, not independent from $X$, neither $...
5
votes
0
answers
218
views
Distribution of $X'X$ if $X\in\mathbb{R}^{T \times N}$ and $X_i'\sim N(\mu,\sigma^2I_N)$
Let $x_i\in\mathbb{R}^N$ be multivariate normal distributed with mean vector $\mu\in\mathbb{R}^N$ and no correlation: $x_i\sim N(\mu,\sigma^2 I_N)$.
Given $T$ iid samples, define the matrix
$$X:=\left(...
5
votes
0
answers
136
views
Relation between asymptotic relative efficiency for tests and estimators
The asymptotic relative efficiency for unbiased estimators is the limit of the ratio of the variances as the $n\rightarrow \infty$. Is there a relation to asymptotic relative efficiency according to ...
5
votes
0
answers
325
views
Restriction matrix for a VAR
In New Introduction to Multiple Time Series Analysis by Luetkepohl (2005), section 5.2.1, it says that one can specify linear restraints for a VAR, $Y = \beta X + U$, in the form
$$
\operatorname{vec}{...
5
votes
0
answers
679
views
Sum of independent Wishart with same degrees of freedom but different scale matrices
Is there any result showing that a sum of independent Wishart with same degrees of freedom but different scale matrices is a Wishart?
For example, if I have two random variables:
$$
Y \sim W_p(n,\...
5
votes
0
answers
356
views
Derivation of prediction intervals for a normally distributed population with unknown population standard deviation
I have via the ISO standard 16269 found the solution to a problem that I've been working on. Based on a couple of independent samples from a normally distributed population, I would like to determine ...