Questions tagged [aic]

AIC stands for the Akaike Information Criterion, which is one technique used to select the best model from a class of models using a penalized likelihood. A smaller AIC implies a better model.

Filter by
Sorted by
Tagged with
0
votes
0answers
12 views

AICc and AIC to compare OLS and MLE estimate regression

I have a question about AICc or AIC to compare the model, which is estimated by ols and mle method. for example , I have a simulate arima model below. ...
0
votes
0answers
6 views

Is it possible to combine drop1() and AICc model selection (AICcmodavg)?

I am using RStudio for GLMM models. I understand individually what drop1() do and AICc model selection do, but I'm not sure if you can use them together. Can I use drop1() to narrow my models to a ...
0
votes
0answers
22 views

Model/variable selection in binomial models for marble counts?

Thank you in advance for your help with this interesting question that has come up in my research. I changed the problem to bags/marbles to simplify. Problem description There are N bags, each ...
1
vote
0answers
28 views

Compute AIC of Linear Mixed-Effects Model by Hand

The Akaike Information Criterion (AIC) quantifies a model's goodness of fit while also taking into account the model's degrees of freedom. In R, there is a function called ...
13
votes
2answers
950 views

What if there is no true data-generating process?

I've been engaging in a number of forecasting efforts recently, and have rediscovered a well-known truth: That combinations of different forecasts are generally better than the forecasts themselves. ...
0
votes
0answers
10 views

Comparing AIC and QIC (GEE and GLM)

I'm doing an analysis of expression data with 5 timepoints and a binary treatment variable. It is common not to use GEEs of any kind in this field and timepoints are often modeled using model.matrix(),...
0
votes
0answers
12 views

AIC comparison between ETS and ARIMA

I was wondering is it relevant to compare AICc on training set between ARIMA and ETS method.
1
vote
1answer
24 views

VAR model selection: is AIC an appropriate measure given that the sample size changes depending on the number of lags?

Given that the sample size of a VAR (or a similar model: VARX, SVAR etc.) reduces by $1$ for each extra dependent-variable lag that I introduce (since we need to drop the empty rows, or ...
5
votes
0answers
25 views

What is Better for Prediction Error: Covariance Penalties or a Test Set?

I'm reading Computer Age Statistical Inference by Efron and Hastie, two statisticians I have a lot of respect for. Section 12.3 discusses Mallows' $C_{p}$, Akaike's information criteria (AIC), and ...
3
votes
1answer
40 views

In choosing best GLM model, what is the significance of the p-value in the model with the lowest AIC?

I am running a glm model in R, with a dataset with a large number of variables (around 25). I have checked for collinearity, and there is quite some between some of groups of variables, so I have run ...
5
votes
2answers
262 views

How to interpret negative values for -2LL, AIC, and BIC?

From a statistical point of view, can -2LL, AIC, and BIC in the table of information criteria from SPSS output be less than zero, ie negative? In this case, how should they be interpreted? Note: These ...
0
votes
1answer
23 views

Using AIC vs Likelihood Ratio test for comparing Lognormal and Powerlaw distributions

I am interested in comparing whether a lognormal or a power law are a better fit for a given set of data. Both distributions have been fit using MLE, with $x_{min}$ determined using KS-minimization a ...
1
vote
1answer
22 views

VAR - VARX model selection

Suppose I have three stationary economic time series $y_i$ that are not cointegrated and I want to investigate the relationship between them. I happen to be unsure about the "endogenousness" ...
0
votes
0answers
14 views

Choosing variables for a semiparametric model

I am trying to create a semiparametric model for university (we were told it HAS to be semiparametric) and I have 11 response variables, some of them categorical and the rest continuous. In the simple ...
0
votes
0answers
15 views

Lag choice for a VAR model

There is something confusing (for me) about this question. I'm working on a VAR model (7 time series) where i've checked for Granger causality (yes) and stationarity (yes). Now, according to the AIC ...
1
vote
2answers
77 views

Removing variables from a linear regression improves $R^2_{adj.}$

I am working on a linear regression model. The complete model with 11 variables in total has a quite low adjusted R-squared ($R^2_{adj.}$) of 0.11. 4 variables have a significant influence on the DV....
1
vote
1answer
41 views

The BIC is different in cph and coxph for the same model

I constructed the same cox regression models by using cph in rms package and coxph in R, but when I compared the two models with BIC, I got 4086.559 for coxph, 4114.43 for cph. But the two models both ...
0
votes
0answers
21 views

How to interpret the effect of a predictor, if the inclusion of time renders the effect insignificant but AIC of the simple model is better?

I research the recurrence of civil war in a sample of 18 countries that experienced civil war with annual longitudinal data over up to 29 years into the postwar period (for some countries less, as war ...
2
votes
1answer
33 views

LASSO vs AIC for submodel selection via nonzero coefficient variable selection

Suppose you have a linear model which you believe has too many variables -- a cubic in 10 lags, for example. You believe, without being certain, that it is probably quadratic, and maybe linear, and ...
2
votes
0answers
10 views

Calculting logLikelihood and AICs "by hand"

I'm a total newbee, when it comes to R and I need some help - my task is to calculate the logLikelihood and the AIC for a logistic regression, without using the logLik() fuction or the AIC() function. ...
1
vote
1answer
27 views

ets() function does not minimize AICc?

I have a question that is similar to this question: ETS function in forecast package is not choosing minimized AICc I see what the author of that question misunderstood but I basically have a reverse ...
0
votes
0answers
15 views

AIC difference after adding offset in negative binomial in R

I'm quite confused about how R calculate likelihood/AIC in glm.nb after adding the offset. In this example, the regression coefficients are the same but the AICs ...
0
votes
1answer
26 views

Why is the AIC score doubled and does it matter?

I have used AIC to do model selection before by just following the classic formula: AIC=2k-2L But as far as I understand the absolute value of this score doesn't matter, only the relative score ...
0
votes
0answers
12 views

problems computing c-hat with a binomial distribution for a glm (AICcmodavg)

I am trying to run a model comparison for generalised linear models in R using the AICcmodavg package. This is one of my models: ...
0
votes
0answers
12 views

intepreting AIC and drop1 of models

Consider the following case: we have continuous response A, and indicator B,C,D. ...
0
votes
0answers
17 views

Mixed-models and PROC MI

I am running into a problem that I cannot seem to figure out. I am fitting a mixed-models logistic regression using PROC GLIMMIX using complex survey data from the Behavioral Risk Factor Surveillance ...
0
votes
0answers
55 views

Linear model selection (AIC/dredge function) with non parametric data (residuals are autocorrelated)

I used the dredge function with MuMIn & lme4 package to do a linear model selection with AIC principle. I have about 10 ...
1
vote
1answer
38 views

Is BIC asymptotically efficient for minimizing prediction error if the true model is being considered?

If a set of models is being compared using BIC and AIC, given the fact that the true model (the one which generated the data) is in this set (and given the other assumptions that guarantee BIC ...
1
vote
1answer
47 views

Information Criteria and Sample size

In the estimation of the univariate time series model, we need to determine the correct order. For the general ARMA (p,q) model, we can determine the true order by information criteria $$AIC(p,q) = ln(...
10
votes
2answers
671 views

Why is AIC not reported with a confidence interval?

In parameter estimation, it's common to report a 95% CI around each parameter. Why don't I see AIC (or deltaAIC) with a CI? If I bootstrap the fitting of two potential models, and get a deltaAIC for ...
0
votes
0answers
19 views

Interprete AIC as significance test and determine significance level for nested models

Let's say we have two nested models. The smaller one (corresponds to $H_0$) has $p_0$ parameters, so its AIC is given by $$AIC_0=-2\log L_0+2p_0$$ The larger model has $p_1:=p_0+d$ parameters of which ...
0
votes
0answers
24 views

AIC and n - original sample size or data points?

I've got a dataset consisting of 5,000 women, who are followed for 28 days, and for each day they have an estimated survival rate. I have taken these 28 survival rates, tried different models, and are ...
0
votes
1answer
47 views

Calculating AIC for a linear regression model

I'm seeing some "inconsistencies" on how R calculates the Akaike Information Criterion (AIC) for linear regression models. I'd like to get its expression so I can calculate it myself. The ...
3
votes
1answer
37 views

Can I use AIC for path models and non-path models?

I am testing competing hypotheses where one hypothesis contains a mediation effects that can be modeled using a path model. The other hypothesis does not include a mediation effect and therefore can ...
1
vote
1answer
167 views

Calculating AIC & BIC

I have an output from two LMER-models and I'd like to calculate AIC & BIC. I believe I've understood the tables correctly, but I'm uncertain regarding the k parameter; have I understood it ...
17
votes
4answers
859 views

Why does the Akaike Information Criterion (AIC) sometimes favor an overfitted model?

As an exercise to develop practical experience working with model selection criteria, I computed fits of the highway mpg vs. engine displacement data from the tidyverse mpg example data set using ...
0
votes
1answer
56 views

Reducing a linear regression (OLS) model by dropping non-significant coefficients

Would it be proper for me to reduce a model by iterating though the coefficients and dropping the ones with high p-values and then refitting and doing this again until all coefficients are significant?...
1
vote
0answers
20 views

Log Likelihood for Two Stage Least Squares (for AIC or BIC)

I'm looking to use a number of information criteria (BIC, AIC, etc.) for Two Stage Least Squares. Of course all information criteria need a log likelihood - and I'm also aware that for a Log-...
7
votes
2answers
766 views

Why is AIC or BIC commonly used in model selections for time series forecasting?

On scikit-learn documentation, I found the following comments about AIC: Information-criterion based model selection is very fast, but it relies on a proper estimation of degrees of freedom, are ...
0
votes
0answers
18 views

GARCH simulation in-sample results make no sense

Idea is to simulate two different garch models with different parameters and later to take loglikelihood of separate models as well as of the concatenation. I wanted to make working example of the ...
1
vote
2answers
81 views

How does AIC vs. LASSO work?

I understand that LASSO and AIC are striking for a balance between model fit and size. However, how do they respectively measure the size/complexity of the model? Does AIC measure the number of ...
2
votes
1answer
34 views

R: Fit Linear Mixed-Effects Model with GAM (mgcv)

Imagine a linear mixed effects model with one random intercept: library(lme4) LMM1 <- lmer(response ~ experience + (1|subject), REML=FALSE, data=train) I'd like ...
1
vote
1answer
23 views

What is the theoretical basis behind counting dispersion as an implicit additional free parameter in R's logLik() function?

Recently I was attempting to calculate the Akaike Information Criterion (AIC) for a set of fitted models in R, using standard packages available on CRAN. A package that I was using seemed to be ...
1
vote
1answer
40 views

How to optimise penalty parameter in ridge regression using AIC

So I know for a ridge regression model, we need to find an optimal $\lambda$ value. I also know that we can achieve this by finding an optimal AIC value, that is, we find the $\lambda$ value that ...
0
votes
1answer
59 views

Does the column ordering matter in the stepwise algorithms used by R?

Suppose I have a large data set with variables $x_1, x_2, \ldots, x_p$ to predict response $y$ where $p$ is very large (however $n >> p$). I would like to perform forward stepwise regression on ...
0
votes
0answers
12 views

When AIC chooses 2 lags, but 2nd lag is insignificant, do I drop it?

If the AIC criterion chose 2 lags, but the 2nd lag is not significant (see p-value), then am I supposed to drop the 2nd lag or leave it?
7
votes
1answer
107 views

GAM: Do shrinkage smooth splines also address for concurvity?

I have a gam model with automatic predictor selection based on cubic splines (bs = cr) and SELECT == T or shrinkage cubic splines (bs = cs) and SELECT == F. Now I'm wondering if predictors affected ...
0
votes
1answer
22 views

How to use the AIC/BIC for overfitting (information criteria) ARIMA

So I am using STATA, I have the log likelihood, AIC and BIC as such: AIC: -112.1838 BIC: -100.2412 log likelihood: 64.23 N= 200 observations So how do I conclude that there is no "over fitting&...
1
vote
1answer
37 views

Negative Log Likelihood for AIC

I was looking at AIC, which is given by AIC = 2K - ln(L). However, to my understanding, and observation, L = Log-Likelihood can be negative. So in the case where L is negative, is AIC not applicable ? ...
2
votes
2answers
116 views

Which statistical test to compare same model with different parameters?

I have two datasets on people buying apples based on weight and price. One dataset in 2019 the other in 2020. I estimate a logit model with Utility = betaWeight * weight + betaPrice * price. Training ...

1
2 3 4 5
19