Skip to main content
Cross Validated

Questions tagged [identifiability]

Ask Question

A model is identifiable if a single set of parameters can be found that will yield the best fit.

264 questions
Newest Active Bountied Unanswered
Filter by
Sorted by
Tagged with
6 votes
2 answers
184 views

I am trying to estimate the values of three unknown parameters in the Ordinary Differential Equation (ODE) using pyMC. Physically, A should be smaller than both B and C. But, given the data and ODE I ...
Auberron's user avatar
  • 183
7 votes
3 answers
366 views

I am dealing with a situation, where my distribution, becomes non-identifiable with respect to the parameters. Can anyone suggest some reliable sources or references wherein I can find the asymptotic ...
Ronik's user avatar
  • 81
5 votes
1 answer
175 views

I was fitting the following binary glmm with glmer: ...
Asier's user avatar
  • 163
0 votes
0 answers
82 views

Let $(Y_j,X_j)_{j=1}^n$ be i.i.d. observations. Consider the hierarchical model \begin{align} Y_j \mid \theta_j &\sim P(\theta_j), \label{eq:obs}\\ E(Y_j) &=\theta_j = g(Z_j), \label{...
Odelia's user avatar
  • 11
0 votes
0 answers
61 views

I want to make a model which predicts y based on previous y and x over time. There are 3 choices in my mind: Model 1: $y_t = \beta_0 + \beta_1 y_{t-1} + \gamma x_{t-1} + \epsilon_t$ where $\epsilon_t ...
user avatar
0 votes
0 answers
109 views

This is a follow up from my previous question Why is the profile likelihood used to determine if a model has a unique solution? . I am interested in knowing how Profile Likelihood can be used to ...
user avatar
1 vote
0 answers
96 views

This is an identification problem of a mixture model: So I have observed data $\{Y_1,...Y_J,X\}$, but full data is $\{Y_1,...Y_J, Z, X\}$, where $\{Y_1,...Y_J\}$ are iid draws from $f_{Y|Z}(y,z)$. $f_{...
Odelia's user avatar
  • 11
3 votes
2 answers
114 views

I'm studying Brady Neal's book. I am doing the exercise on page 60, chapter 6, active reading exercise. ($T$ = treatment, $Y$ = target). I am trying to figure out in the above graphs assuming the ...
yo wa's user avatar
  • 169
0 votes
0 answers
41 views

I’m fitting a mixture model in Stan to strictly positive data. Each component distribution has a mean that can be expressed by some summation of a and b. (a and b are two “factor” parameters shared ...
brightlilac's user avatar
  • 31
7 votes
1 answer
311 views

Consider the following model: $X_i \mid N \sim \text{Binomial}(N, \theta)$ and $N \sim \text{Neg-Bin}(\mu, \varphi)$ (parametrized by mean and dispersion). Is this model identifiable? How to prove it? ...
Ezequiel Santos's user avatar
  • 71
6 votes
2 answers
875 views

I am trying to find some references which explain how do we know that a statistical model is identifiable. My current understanding of identifiability is when only a single set of solutions exist for ...
stats_noob's user avatar
  • 1
1 vote
1 answer
76 views

I'm currently using Structural vector autoregressive models by Kevin Kotzé to learn Vector Autoregression. One of the points that it makes is the following: the number of restrictions that we need to ...
stats_learner's user avatar
  • 43
4 votes
1 answer
330 views

If my regression model $$ y_i = \alpha + \beta x_i + \epsilon_i $$ suffers from OVB the error contains one variable which we assume correlated with $$ \epsilon_i = \gamma w_i + u_i $$ my estimate of $\...
Three Diag's user avatar
  • 517
1 vote
1 answer
107 views

Is there any connection between multicollinearity and problem of identification in Simultaneous Equations Model? I know multicollinearity is the occurrence of high intercorrelations among two or more ...
CrunchySia24's user avatar
  • 11
3 votes
1 answer
96 views

I have this expression $$\begin{multline} p( Y \mid \text{do}(Z=z)) = \\ \int_{B, S, W, X} dBdSdWdX \ \ P(B | S) P(W | B, S) P(X | B, S, Z=z) \\ \left[ \int_{Z'} dZ' P(Z'| B,S,W) P(Y | B, S, W, X, Z'...
Astrid's user avatar
  • 1,029

15 30 50 per page
1
2 3 4 5
…
18