EconBase
← Back to paper

Asymptotic in a class of network models with an increasing sub-Gamma degree sequence

Extracted main text — title through conclusion, appendix excluded. This is what our citation measures are computed over, published so the extraction can be checked by eye.

34,063 characters · 9 sections · 39 citation commands

Rendered from LaTeX for readability, not typeset faithfully. Citation keys are highlighted; maths is left as source; figures, tables and equation environments are summarised rather than reproduced; unrecognised commands are greyed out so nothing is silently dropped. Email addresses are removed.
center[center omitted — 873 chars of source]

\footnotetext[3]{$^*$Correspondence author. Email: [email removed] (Haoyu Wei). Luo's research is partially supported by the Fundamental Research Funds for the Central Universities of South-Central Minzu University(Grant Number:CZQ22003) and by National Statistical Science Research of China (2022LY051) and by the Open Research Fund of Key Laboratory of Nonlinear Analysis & Applications (Central China Normal University), Ministry of Education, P.R. China. \\ Jing Luo and Haoyu Wei are co-first authors. Email:[email removed] (Jing Luo), {[email removed]} (Xiaoyu Lei), [email removed] (Jiaxin Guo).} \vskip 1.5mm

abstractFor the differential privacy under the sub-Gamma noise, we derive the asymptotic properties of a class of network models with binary values with a general link function. In this paper, we release the degree sequences of the binary networks under a general noisy mechanism with the discrete Laplace mechanism as a special case. We establish the asymptotic result including both consistency and asymptotically normality of the parameter estimator when the number of parameters goes to infinity in a class of network models. Simulations and a real data example are provided to illustrate asymptotic results. \vskip 5 pt Key words: Consistency and asymptotic normality; Network data; Differential privacy; sub-Gamma degree sequence\\ { \bf Mathematics Subject Classification:} 62E20, 62F12.

\vskip 5 pt

Introduction

In network data analysis, the privacy of network data has received wide attention as the disclosed data is likely to contain sensitive information about individuals and their social relationships (sexual relationships, email exchanges, money transfers, etc). Analyzing this type of data can uncover valuable information that can be used to address many important social concerns, such as disease transmission, fraud detection, precision marketing, and more. In recent years, the urgent need to solve the problem of network privacy protection has led to the rapid development of algorithms for securely publishing network data or aggregating network data (see Zhou2008A, Yuan2011Personalized, Cutillo2010Privacy, and Lu2014Exponential). However, the unstructured characteristics of network data bring great challenges to the statistical inference (see fienberg2012brief, Mosler2015) . Data privacy protection is usually achieved by adding some noise to the data. The most typical method is differential privacy[Dwork2006]. Due to the unstructured characteristics of network data and the data with noise, there are relatively few theoretical studies on statistical asymptotic behavior of noise-based network data.

The Erd\"{o}s-R\'{e}nyi model (see Erd1959On) is generally acknowledged as one of the earliest random binary graph models, in which each edge occurs with the same probability independent of any other edge. However, it lacks the ability to capture the extent of degree heterogeneity commonly associated with network data in practice. To capture the heterozygous of nodes' degree, a class of models have come into use for analyzing binary undirected networks network data. The simplest network model is the binary network model, which mainly uses the degree sequence to capture the real network [albert2002]. The $\beta$-model has undirected binary weighted, which is known for binary arrays whose distribution only depends on the row and column totals (see degree1britton2006generating,degree3bickel2011method,degree4zhao2012consistency,hillar2013maximum). When the number of network nodes tends to infinity, many literature have studied the Maximum Likelihood Estimator (MLE) of the $\beta$-model(see chatterjee2011random; degree2blitzstein2011sequential; rinaldo2013maximum). The asymptotic normality of the MLE of the $\beta$-model is further studied by yanxu2013. Fan2002Connected proposes another type of binary network model. This model is called log-linear model, where the edge probability $p_{ij}$ between vertices $i$ and $j$ is $w_{i}w_{j}/(\sum_{k=1}^{n} w_{k})$ under the normalization constraint $w_{i}^{2}\leq (\sum_{k=1}^{n} w_{k}) $($i=1,...,n$), where $w_{i}$ is referred to as the weight of vertex $i$. Moreover, Olhede and Wolfe (2012) obtained the approximate value of the parameter estimation of a class of binary network in the case of limited sample network nodes. Until recent years, Karwa2016 has given research on the asymptotic theory of adding discrete Laplace noise to network data. However, when the general noise is added to a class of this network model, the asymptotic theory of parameter estimators is still unknown.

In this paper, we mainly study the asymptotic theory of a class of network models with sub-Gamma degree sequences. It is different from the method of adding noise to network data in Karwa2016, luo2022asymptotic and luo2021Ordered. Therefore, they only considered the asymptotic theory of the parameters for the $\beta-$model in the case of Laplace noise. Here, we consider the general distribution of noise variables, and Laplace distribution is only a special case. And then, our method does not need to deal with the degree sequence of noise, so we can directly use the degree sequence with noise to study the statistical inference of the network model. Furthermore, the probability-mass or density function of the edge $a_{ij}$ only depends on the sum of $\alpha_i^{*}$ and $\alpha_j^{*}$, where $\alpha_i^{*}$ denotes the strength parameter of vertex $i$. These are the main contributions of this paper.

For the rest of this article, we state as follows. In Section 2, we give null models for undirected network data and sub-Gamma degree sequences. In Section 3, we give a uniform asymptotic result. In Section 4, we illustrate several applications of our main results. Summary and discussion are given in Section 5. Proofs are given in the Appendix.

Notation: For $0< q\leq\infty$, we write $\| \beta\|_{q}:=(\sum_{i=1}^{p}|\beta_{i}|^{q})^{1/q}$ as the $\ell_{q}$-norm of a $p-$dimensional vector $\beta$. If $q=\infty$, we have $\| \beta\|_{\infty}:=\max\limits_{i=1,...,p}|\beta_{i}|$. For a subset $C\subset \mathbb{R}^n$, let $C^0$ and $\overline{C}$ denote the interior and closure of $C$ in $\mathbb{R}^n$, respectively. Let $\Omega(\mathbf{x}, r)$ denote the open ball $\{\mathbf{y}: \|\mathbf{x}-\mathbf{y}\|< r \}$, and $\overline{\Omega(\mathbf{x}, r)}$ be its closure.

Null models for undirected network data

Several Models

Following yanxu2013, we consider an undirected graph $\mathcal{G}_{n}$ on n ( $n \geq 2$ ) agents labeled by $1,...,n$. Let $a_{ij}\in\{0,1\}$ be the weight of the undirected edge between $i$ and $j$. That is, if there is a link between $i$ and $j$, then $a_{ij}=1$; otherwise, $a_{ij}=0$. Denote $A=(a_{ij})_{n \times n}$ as the symmetric adjacency matrix of $\mathcal{G}_{n}$. We assume that there are no self-loops, i.e., $a_{ii}=0$. Define $d_{i}=\sum_{j\neq i}^{n}a_{ij}$ as the degree of vertex $i$ and $\mathbf{d}=(d_{1},\dots,d_{n})^{\top}$ as the degree sequence of the graph $\mathcal{G}$. We suppose that the adjacency matrix $A$ has independent Bernoulli elements such that $\mathrm{P} (a_{ij}=1) = p_{ij}$ and specify the corresponding family of probability null models for $A$. Here $\boldsymbol{\alpha}^{*}=(\alpha_{1}^{*},\alpha_{2}^{*},...,\alpha_{n}^{*})^{\top}$ is a parameter vector. The parameter $\alpha_i^*$ quantifies the effect of the vertex $i$. To this end, let $\varepsilon(\cdot,\cdot): \mathbb{R}^2 \mapsto \mathbb{R}$ be a smooth bivariate function satisfying $ \varepsilon(x,y) = \varepsilon(y,x)$. Consider the model $\emph{M}_{\varepsilon}$ specified $p_{ij}$ as

equation[equation omitted — 141 chars of source]

so that we obtain a class of log-linear models indexed by $\varepsilon$. In fact, this class encompasses three common choices of link functions:

equation[equation omitted — 94 chars of source]
equation[equation omitted — 187 chars of source]
equation[equation omitted — 109 chars of source]

To see this, set $\varepsilon(\alpha_{i}^{*},\alpha_{j}^{*}) = -\log\{1+ \exp(\alpha_{i}^{*}+ \alpha_{j}^{*})\}$ for the model $\emph{M}_{logit}$. As we have seen, the logit-link model $\emph{M}_{logit}$ is an undirected version of Holland1981An exponential family random graph model without reciprocal parameter. As noted by yanxu2013, the degree sequence of $\mathcal{G}$ is sufficient for $\alpha$ in this case, and they derived its asymptotic normality. The log-link model $\emph{M}_{\log}$ can be considered as an undirected version of expected degree model with $\varepsilon(\alpha_{i}^{*},\alpha_{i}^{*}) = 0$ constructed by Fan2002Connected. Setting $\varepsilon(\alpha_{i}^{*},\alpha_{j}^{*}) = \log\{1- \exp(\alpha_{i}^{*}+ \alpha_{j}^{*})\}-(\alpha_{i}^{*}+ \alpha_{j}^{*})$, we get the complementary log-log link model $\emph{M}_{\rm cloglog}$ (see Mccullagh1989Generalized).

The sub-Exponential/-Gamma noisy sequence

In this section, we will prepare the probability and distribution preliminary for later network analysis. This section can be divided into two parts. In the first part, we are going to recap the definition of a specific type of distributions and state their basic properties.

definition[Sub-Gaussian distribution] A random variable $X \in \mathbb{R}$ with mean zero is sub-Gaussian with variance proxy $\sigma^{2}$ if its MGF satisfies $$\mathrm{E}[\exp (s X)] \leq \exp \left(\frac{\sigma^{2} s^{2}}{2}\right), \quad \forall s \in \mathbb{R}.$$ In this case we write $X \sim \operatorname{subG}\left(\sigma^{2}\right) .$

Similarly, we define that a random variable is called sub-exponential if its survival function is bounded by that of a particular exponential distribution.

definition[Sub-exponential distribution] A random variable $X \in \mathbb{R}$ with mean zero is sub-exponential with parameter $\sigma^{2}$ (denoted by $X \sim \operatorname{subE}(\lambda))$ if its MGF satisfies \begin{equation} {\mathrm{{E}}}e^{s X} \leq e^{\frac{s^{2}\lambda ^{2}}{2}} \quad for all |s| <\frac{1}{\lambda}. \end{equation}

Obviously, sub-Gaussian random variables are sub-exponential but not vice verse. And there are some equivalent definitions of sub-exponential distributions, see Rigollet2019, which can lead to sub-exponential norm used in concentration inequalities. The equivalent definitions and other details can be seen in the Appendix.

definition[Sub-exponential norm] The sub-exponential norm of $X$, denoted $\|X\|_{\psi_1}$, is defined as \begin{equation} \|X\|_{\psi_1} = \inf \left\{ t>0 :\; \mathrm{E} \exp(|X|/t) \le 2 \right\}. \end{equation}

The above norm provides us with a useful tool to connect MGF and the defined norm, and hence makes it possible to give concentrations for the sub-exponential variables. The following lemma confirms Definition (ref) would give a concise form of concentrations.

lemma[Properties of sub-exponential norm] If $\mathrm{E} \exp(|X|/\|X\|_{{\psi _1}} ) \le 2$, then we have \begin{enumerate}[\rm{(}a\rm{)}] • Tail bounds \begin{equation} \mathrm{P} \{ |X| > t \} \leq 2\exp(-t/\|X\|_{\psi _{1}}) \quad for all t \geq 0; \end{equation} • Moment bounds $$\mathrm{E} |X|^k \le 2{\|X\|_{{\psi _1}}^k}k! \quad \text{for all integer}~k \ge 1;$$ • If $\mathrm{E} X=0$, we get the MGF bounds \begin{equation} \mathrm{E} e^{s X} \leq e^{({\rm{2}} \left\| X \right\|_{{\psi _1}})^{2}s^{2}} for all |{s}|<1/(2\left\| X \right\|_{{\psi _1}}) \end{equation} which gives $X \sim \operatorname{subE}(2\left\| X \right\|_{{\psi _1}})$. \end{enumerate}

Lemma (ref)(c) implies that the following user-friendly concentration inequality would contain all known constant. One should note that Theorem 2.8.1 of Vershynin2018 includes an unspecific constant, so it is inefficacious when constructing non-asymptotic confident interval for sub-exponential sample mean.

corollary[Concentration for sub-exponential sum of r.v.s, zhang2020concentration] Let $\{ X_{i}\} _{i = 1}^n $ be zero mean independent sub-exponential distribution with $\|X_i\|_{\psi_1}\le \infty$. Then for every $t \ge 0$, \[\mathrm{P} ( {| {\sum\limits_{i = 1}^n {{X_i}} }| \ge t} ) \le 2 \exp\{ { - \frac{1}{4}( {\frac{{{t^2}}}{\sum\nolimits_{i = 1}^n {2\|X_i\|_{\psi_1}^2}} } \wedge \frac{t}{\mathop {\max }\limits_{1 \le i \le n}\|X_i\|_{\psi_1}}}) \}.\]

Hay2009 and Karwa2016 used the Laplace mechanism to provide privacy protection in which independent and identically distributed Laplace random variables are added into the input data. Here, we consider a general distribution for the noisy variables with the Laplace distribution as a special case.

example[Laplace r.vs] {\color{black}{ A r.v. $X$ follows a Laplace distribution (${{\textrm {Laplace}}(\mu ,b)},~\mu \in \mathbb{R} ,b>0$) if its probability density function is $f(x )=\frac{1}{2 b} e^{-\frac{|x-\mu|}{b}}$. The Laplace distribution is a distribution of the difference of two independent identical exponential distributed r.vs, thus it is also sub-exponential distributed by using Proposition (ref)(a). The graph of $f(x)$ are like two exponential distributions which are spliced together back-to-back.}}
example[Geometric distributions] {\color{black}{The geometric distribution $X\sim \mathrm{Geo} (q) $ for r.v. $X$ is given by $\mathrm{P} (X=k) ={(1 - q)} {q^{k-1}},~~q \in (0,1),k=1,2,\cdots.$ The mean and variance of ${\rm{Geo}} (q) $ are $\frac{{1 - q}}{{q}}$ and $\frac{{1 - q}}{{{{q}^2}}}$ respectively. Apply Lemma 4.3 in hillar2013maximum, we have ${({\rm{E}}|X{|^k})^{1/k}} < \frac{{2k}}{{ - \log (1 - q)}}$. Followed by the triangle inequality applied to the $p$ -norm and Jensen's inequality for $k \ge 1$, we have $$ \mathrm{E}[|X-\mathrm{E}X|^{k}]^{1 / k} \le \mathrm{E}[|X|^{k}]^{1 / k}+|\mathrm{E}[X]| \leq 2 \mathrm{E}[|X|^{k}]^{1 / k}\le \frac{{4k}}{{ - \log (1 - q)}}. $$ Then Proposition (ref)(3) shows that the centralized geometric distribution is sub-exponential with $K_3=\frac{{4}}{{ - \log (1 - q)}}$. }}
example[Discrete Laplace r.vs] A r.v. $X$ obeys the discrete Laplace distribution with parameter $p \in(0,1),$ denoted by $\mathrm{DL}(p),$ if \[ f_{p}(k)=\mathrm{P}(X=k)=\frac{1-p}{1+p} p^{|k|}, \quad k \in \mathbb{Z}=\{0,\pm 1,\pm 2, \ldots\}. \] Similar to Laplace distribution, the discrete Laplace r.v. is the difference of two independent identical geometric distributed r.vs (see Proposition 3.1 in Inusah2006A). Since geometric distribution is sub-exponential in the previous example, the Proposition (ref)(a) implies that discrete Laplace is also sub-exponential distributed. In differential privacy of network models, the noises are assumed from the discrete Laplace distribution (see fan2020 and references therein).

In statistical applications, we sometimes do not expect the bounded assumption in Hoeffding's inequality, the following Bernstein's inequality for a sum of independent random variables allows us to estimate the tail probability by a weaker version of exponential condition on the growth of the $k$-moment (like a condition of the exponential MGF) without any assumption of boundedness.

lemma[Bernstein's inequality] The centred independent random variables $X_1,\ldots, X_n$ satisfy the growth of moments condition \begin{equation} {\rm{E}}{\left| {{X_i}} \right|^k} \le \frac{1}{2}\nu_i^2{\kappa_i^{k - 2}}k!, (i = 1,2, \cdots ,n), for all k\ge 2 \end{equation} where $\{\kappa_i\}_{i=1}^n$ and $ \{\nu _i\}_{i=1}^n$ are constants independent of $k$. Denote ${\nu _n^2} = \sum\limits_{i = 1}^n {v _i^2} $ (the fluctuation of sums) and $\kappa= \mathop {\max }\limits_{1 \le i \le n} \kappa_i$. Then we have $ {\rm{E}} e^{s X_{i}} \leq e^{s^{2} \nu_{i}^{2} /(2-2 \kappa_i|s|)}. $ And for $t>0$ \begin{equation} \mathrm{P} \left( {\left| {{S_n}} \right| \ge t} \right) \le 2\exp \left( { - \frac{{{r^2}}}{{2{\nu _n^2} + 2\kappa t}}} \right), \mathrm{P}( {\left| {{S_n}} \right| \ge \sqrt {2\nu _n^2t} + \kappa r}) \le 2{e^{ - t}}. \end{equation}

The proof Bernstein's inequality for the sum of independent random variables can be founded in p119 of Gin2015Mathematical.

Like the sub-Gaussian, boucheron2013concentration defines the sub-Gamma r.v. based on the right tail and left tail with variance factor $v$ and scale factor $b$.

definition[Sub-Gamma r.v.] A centralized r.v. $X$ is {\em sub-Gamma} distributed with variance factor $\upsilon>0$ and scale parameter $c>0$ (denoted by $X \sim \mathrm{sub}\Gamma(\upsilon,c)$) if \begin{equation} \log (\mathrm{{E}}{e^{s{X}}}) \leq \frac{s^{2}}{2} \frac{\upsilon}{1-c|s|}, \quad \forall 0<|s|<c^{-1}. \end{equation}

The sub-exponential moment condition (ref) would imply that Bernstein's moment condition (ref) is observed as $$\log (\mathrm{{E}}{e^{s{X}}}) \le {\frac{s^{2}\lambda ^{2}}{2}} \le \frac{s^{2}\lambda ^{2}}{ {2(1 - \lambda|s|)}},~\forall~|s| <\frac{1}{\lambda}.$$

lemma[Concentration for sub-Gamma sum, Section 2.4 of boucheron2013concentration] Let $\{ X_{i}\} _{i = 1}^n $ be independent $\{\mathrm{sub}\Gamma(\upsilon_i,c_i)\} _{i = 1}^n$ distributed with zero mean. Define $c= \mathop {\max }\limits_{1 \le i \le n} c_i$, then \begin{enumerate}[\rm{(}a\rm{)}] • Closed under addition: $ S_n:=\sum\limits_{i = 1}^n {{X_i}} \sim {\rm{sub}}\Gamma({\sum\limits_{i = 1}^n\upsilon_i},c)$; • $\mathrm{P} (|S_n|\geq t)\leq 2 \exp \left(-{\frac {t^{2}/2}{\sum _{i=1}^{n}\upsilon_i+ct}}\right)$ and $\mathrm{P} \{|S_n|>(2t \sum_{i=1}^{n}\upsilon_i)^{1/2}+c t\} \leq 2 e^{-t},~\forall~t \ge 0$; • If $X \sim \mathrm{sub}\Gamma(\upsilon,c)$, the even moments bounds satisfy $\mathrm{E}{X^{2k}} \le k!{({8v} )^k}+(2k)!{({4c} )^{2k}},~k\ge 1.$ \end{enumerate}

The concentration inequalities introduced in above only concerns the linear combinations of independent random variables. For lots of applications in high-dimensional statistics, we have to control the maximum of the n r.vs when deriving error bounds for the proposed estimator. In our proof of Theorem (ref), the following maximum inequality is crucial.

lemma[Concentration for maximum of sub-Gamma random variables] Let $\{ X_{i}\} _{i = 1}^n $ be independent $\{\mathrm{sub}\Gamma(\upsilon_i,c_i)\} _{i = 1}^n$ distributed with zero mean. Denote $\max\limits_{i=1,...,n}\upsilon_{i} = \upsilon$ and $\max\limits_{i=1,...,n}c_{i} = c$, we have \begin{equation} \mathrm{E} (\max_{i=1,\ldots,n}|X_i|) \leq \sqrt{2\upsilon \log (2n)} + c \log (2n). \end{equation}

Estimation and its asymptotic properties

In this section, we will derive the asymptotic results for the estimator with an increasing sub-Gamma degree sequence. Note that $\mathrm{E}(a_{ij})$ only depends on the $e^{\alpha_{i}^{*}+ \alpha_{j}^{*}+\varepsilon_{ij}(\alpha_{i}^{*}+\alpha_{j}^{*})}$. Let $\mathbf{d}=(d_{1}, \ldots, d_{n})^{\top}$ be the degree sequence of graph $\mathcal{G}_{n}$. We assume that random variables $\{ e_{i}\}_{i=1}^n$ are mutually independent and distributed in sub-gamma distributions $\{{\rm{sub}}\Gamma(\upsilon_i,c_i)\} _{i = 1}^n$ with respective parameters $\{(\upsilon_i,c_i)\} _{i = 1}^n$. Then we observe the noisy sequence $\tilde{d}$ instead of $d$, where

equation[equation omitted — 120 chars of source]

We use moment equations to estimate the degree parameter with the noisy sequence $\tilde{d}$ instead of $d$. Define a system of functions:

gather*[gather* omitted — 311 chars of source]

Now, we define our estimator $\widehat{\boldsymbol{\alpha}}$ as the solution to the equation $F(\boldsymbol{\alpha})=0$, i.e.,

equation[equation omitted — 126 chars of source]

It is not hard to see that the estimator is actually induced by the moment equation $\mathbf{\tilde{d}} = \mathrm{E} (\mathbf{d})$.

These asymptotic results of $\widehat{\boldsymbol{\alpha}}$ hold for all $\varepsilon$ satisfying the following condition.

assumptionFor all pairs of node $i$ and node $j$, all choices of $k,l$ and $m$ $(k,l,m=1,...,n)$, the function $\varepsilon_{i,j}$, $\partial \varepsilon_{i,j} / \partial \alpha_{k} $, $\partial \varepsilon^{2}_{i,j} / \partial \alpha_{k} \partial \alpha_{l} $, and $\partial \varepsilon^{3}_{i,j} / \partial \alpha_{k}^{*} \partial \alpha_{l} \partial \alpha_{m} $, are sub-exponential in $\alpha_{i} +\alpha_{j} $. That is, there exists a constant $M_{0}$ such that the absolute values of these functions are bounded by $M_{0} \exp (\alpha_{i} + \alpha_{j})$.

Note that the solution to the equation $F({\boldsymbol{\alpha}})=0$ is precisely the moment estimator. Here, we consider the symmetric parameter space \[D = \{ {\boldsymbol{\alpha}} \in \mathbb{R}^{n}: -Q_{n}\leq \alpha_{i} + \alpha_{j} \leq Q_{n}, \ Q_{n} >0, \ 1\leq i<j \leq n\}.\] The uniform consistency of $\widehat{\boldsymbol{\alpha}}$ and asymptotic distribution of the parameter estimator are stated as follows, and the proofs are given in Appendix.

theorem[Consistency] If \begin{equation*} e^{Q_{n}+e^{Q_{n}}} = \left(\frac{\sqrt{(n-1)\log(n-1)}+\sqrt{2\upsilon \log (2n)} + c \log (2n)}{n} \right)^{-1/34} \end{equation*} and $\max_{i=1,...,n}\upsilon_{i} = \upsilon$, $\max_{i=1,...,n}c_{i} = c$, then as $n \rightarrow \infty$, the estimator $\widehat{\boldsymbol{\alpha}}$ exists and satisfies \begin{equation} \|\boldsymbol{\widehat{\alpha}} - \boldsymbol{\alpha}^{*} \|_\infty = O_p \left( \frac{1}{n}e^{15Q_{n}+e^{Q_{n}}} \big(\sqrt{n \log n }+\sqrt{2\upsilon \log (2n)} + c \log (2n) \big) \right) = o_p(1). \end{equation}

We use the Newton-Kantovorich theorem to prove the consistency of the estimators by constructing the Newton iterative sequence. This technical step is different from Chatterjee et al. (2011) and provides a simple proof. The proof of the theorem is in Appendix.

theorem[Asymptotic normality] Using the same notations as Theorem (ref), if $$\frac{1}{n}e^{45Q_{n}+e^{Q_{n}}} \left(\sqrt{(n-1)\log(n-1)}+\sqrt{2\upsilon \log (2n)} + c\log (2n)\right)^2 = o(n^{1/2})$$ then for any fixed $k\ge 1$, as $n \to\infty$, the vector consisting of the first $k$ elements of $(B^{-1})^{1/2}(\widehat{\boldsymbol{\alpha} }-\boldsymbol{\alpha}^{*} )$ is asymptotically distributed as $N(0,\mathbf{I}_k)$, where $(B^{-1})^{1/2}=\operatorname{diag}(v_{11}^{1/2}, \ldots, v_{nn}^{1/2})$ with \begin{equation*} v_{ii}=\sum_{j \neq i}^{n}e^{\alpha_{i}^{*}+\alpha_{j}^{*}+\varepsilon_{i,j}(\alpha_{i}^{*},\alpha_{j}^{*})}\left(1+\frac{\partial \varepsilon_{ij}(\alpha_{i}^{*},\alpha_{j}^{*}) }{\partial \alpha_{i}^{*} } \right). \end{equation*}

The proof of the theorem is in Appendix.

Numerical studies

In this section, we will evaluate the asymptotic result in Theorems (ref) through numerical simulations and a real data example. The simulation will be conducted using Hermite distributions which we have introduced and proved its properties under the framework of Section 2.

Simulation studies

For the simulation study of all models $M_{\varepsilon}$, we set $\alpha_{i}^{*} = i*L/n$ for $i=1,...,n$. Other parameter settings in simulation studies are listed as follows. We consider three different values $-\log(\log(n))^{1/3}$, $ -\log(\log(n))^{1/2}$, and $-\log(\log(n))$ for $L$ in case of model $M_{log}$. For model $M_{logit}$, we consider three different values $L=0$, $\log(\log(n))$ and $\log(n)^{1/2}$. In case of model $M_{cloglog}$, we set three different values $L=\log(\log(\log(n)))^{3/2}, \log(\log(\log(n)))^{5/4}$ and $\log(\log(\log(n)))$. The noise $e_{i}(i=1,...,n)$ is the difference between two independent and identically distributed Hermite distributions (see Definition (ref)). We consider three cases of Hermite distribution where parameters $a_1=0.01$, $a_2=(\Lambda-0.01)/4, m=2$; $a_1=\Lambda-0.01$, $a_2=0.025$, $m=2$ and $a_1=4*\Lambda/5$, $a_2=\Lambda/5$, $m=2$. Here, we set $\Lambda = 2*\exp(-\lambda_{0}/2)/(1-\exp(-\lambda_{0}/2))^{2}$ and $\lambda_{0}=2$. We consider two values for $n=100$ and $n=200$. Note that by Theorems (ref), $\widehat{\xi}_{ij} =(\widehat{\alpha}_i + \widehat{\alpha}_j-(\alpha_{i}^{*} + \alpha_{j}^{*}))/(1/\widehat{v}_{ii}+1/\widehat{v}_{ii})^{1/2}$ is an asymptotically normal distribution, where $\hat{v}_{ii}$ is the estimator of $v_{ii}$ by replacing $\alpha_i$ with $\hat{\alpha}_i$. The quantile-quantile(QQ) plots of $\xi_{ij}$ are drawn. We ran $10000$ simulations for each scenario.

We simulate with $n=100$, $n=200$, three values for $L$ and three cases of Hermite distribution to find that the QQ-plots for each combination are similar. In order to save the place, we only present the QQ plots of three model in Figure (ref), Figure (ref) and Figure (ref) when $n=200$, $a_1=4*\Lambda/5$, $a_2=\Lambda/5$, and $m=2$ for each case. The horizontal and vertical axes are the theoretical and empirical quantiles respectively, and the straight lines correspond to the reference line $y=x$. In Figure (ref), we first observe that the empirical quantiles agree well with the ones of the standard normality of $\widehat{\xi_{ij}}$, expect for pair $(n/2, n/2+1)$ and $(n-1, n)$ when $L=-\log(\log(n))$ in the case of model $M_{log}$. For the case of model $M_{logit}$, there are no notable derivations from the standard normality for each scenario in Figure (ref). We also observe that the empirical quantiles agree well with the ones of the standard normality of $\widehat{\xi}_{ij}$, expect for pair $(n/2, n/2+1)$ and $(n-1, n)$ when $L=\log(\log(\log(n)))$ in Figure (ref).

The coverage probability of the $95\%$ confidence interval for $\alpha_i-\alpha_j$, the length of the confidence interval and the frequency that the MLE did not exist are reported in Table (ref), Table (ref) and Table (ref). We can see that the length of estimated confidence interval increases as $L$ increases for fixed $n$, and decreases as $n$ increases for fixed $L$.

A data example

We use the KAPFERER TAILOR SHOP network dataset created by Bruce Kapferer which is downloaded from \url{http://vlado.fmf.uni-lj.si/pub/networks/data/ucinet/ucidata.htm}. In each study they obtain measures of social interaction among all actors. In this network data, $``1''$ represents they have a friend relationship between two actors, otherwise, it is denoted as $``0''$. Because the estimate $\hat{\alpha}$ does not exist when the degree is zero, we remove the vertex 17 and 22 whose degree is zero before analysis, so the network with the left 37 vertices in each table remains. When the parameters of three kinds of noise distribution function $e_{i}(i=1,...,n)$ are set, the parameter estimation is similar in case of model $M_{logit}$, $M_{log}$ and $M_{cloglog}$. Thus we here only show the parameter estimation under the case of noise distribution function parameter for $a_1=4*\Lambda/5$, $a_2=\Lambda/5$ and $m=2$. From Figure (ref), we first observe the scatter plots of the noisy degree sequence $\tilde{d}$ corresponding to the parameter estimation $\hat{\alpha}$ under model $M_{\varepsilon}$ and the value of $\hat{\alpha}$ increases as the number of $\tilde{d}$ climbs. The larger the estimated parameters $\hat{\alpha}$, the actors have more friends. For these three cases of model $M_{\varepsilon}$, the estimated parameters and their standard errors as well as the $95\%$ confidence intervals and the size of noisy degree sequences are reported in Table (ref), Table (ref) and Table (ref) respectively. The value of estimated parameters reflects the corresponding size of noisy degrees. For example, the large five degrees are $26,17,16,15,14$ for vertices $16,18,32,11,12$ which also have the top five influence parameters at $0.36,-0.05,-0.12,-0.18,-0.25$. On the other hand, the five vertices with smallest parameters $-2.89,-2.19,-1.79,-1.28,-1.10$ have degrees at $1,2,3,5,6$. In Table (ref) and Table (ref), the larger the parameter $\hat{\alpha}$, the greater the degree $\tilde{d}$ of the node. This is the same as the conclusion in Table (ref).

Summary and discussion

In this paper, we release the degree sequences of the class of binary networks under the sub-Gamma noisy mechanism. We establish the asymptotic result including the consistency and asymptotically normality of the parameter estimator when the number of parameters goes to infinity. By using the Newton-Kantorovich theorem, we try to ignore adding noisy process and obtain the existence and consistency of the parameter estimator satisfying equation $\mathbf{\tilde{d}} = \mathrm{E} (\mathbf{d})$. Furthermore, we give some simulation results to illustrate that the asymptotic normality behaves well under model $\emph{M}_{\log}$, $\emph{M}_{logit}$ and $ \emph{M}_{cloglog}$. However, an edge in networks takes not only binary values but also weighted edges in many scenarios. We will investigate null models for these directed weighted networks in the future. It is worth noting that the conditions imposed on $Q_{n}$,$a_1$, $a_2$, and $m=2$ may not be the most possible. In particular, the conditions guaranteeing the asymptotic normality are stronger than those guaranteeing the consistency. Simulation studies suggest that the conditions on $Q_{n}$ might be relaxed. It can be noted that the asymptotic behavior of the parameter estimator depends not only on $Q_{n}$ $a_1$, $a_2$, and $m=2$, but also on the configuration of all the parameters, We will investigate this in future studies.

In this paper we derived individual parameter asymptotic properties, and we can also study on a linear combination of all the parameter estimation in binary networks with noisy degree sequence in the future work. In our paper, we only consider the model heterogeneity parameter. In network data, the second distinctive feature inherent in most natural networks is the homophily phenomenon. Yan2019 established the uniform consistency and asymptotic normality of the heterogeneity parameter and homophily parameter estimators. On the other hand, sub-Weibull variables, as an extension of sub-Gamma variables, enable variables have heavier tails, which may also be consider in networks models. Fortunately, there have been some articles investigating concentration of sub-Weibull variables, see zhang2021sharper for instance. And we further investigate a central limit theorem for a linear combination of all the maximum likelihood estimators of degree parameter when the number of nodes goes to infinity[luo2020asymptotic]. And the asymptotic theory of the affiliation network model with noise sequence is also worth further study[luo2021affiliation]. We will investigate these aspects in future studies.