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.

learn more… | top users | synonyms

0
votes
1answer
29 views

p-values of the coefficients or AIC for model selection in multiple regression

I´ve got two models from a multiple linear regression (A and B, see below) and don´t know which to select. I want to predict a value called AW as good as possible, so I´d like to have the highest r². ...
3
votes
1answer
42 views

Post Model Selection Inference problems - which remedies exist?

Recently, Hannes Leeb from Yale University and Benedikt Pötscher from the University of Vienna have published a series of papers dealing with what they call Post Model Selection Inference problems.* ...
10
votes
2answers
255 views

Different AIC definitions

From Wikipedia there is a definition of Akaike's Information Criterion (AIC) as $ AIC = 2k -2 \log L $, where $k$ is the number of parameters and $\log L$ is the log-likelihood of the model. However, ...
1
vote
1answer
19 views

Combining AIC Scores

I am trying to combine two AIC scores from different parts of a single data set. I have two sigmoidal models that each predict about half of the data (imagine two sigmoidal curves attached end to ...
0
votes
0answers
26 views

substituting zeros in a Gamma regression

I modeled some right skewed data with a Gamma GLM (log link). This is common practice in my field. However, some observations have a value of zero and the Gamma distribution is only defined on the ...
0
votes
2answers
46 views

AIC/BIC and data transformation

Can you use AIC/BIC to compare models on untransformed data with models on transformed data (such as log, inverse hyperbolic sine, etc.)? I.e. if a model using logged data gives an AIC = 53.62 and a ...
0
votes
0answers
10 views

How to calculate AIC for multiple participants

I have a number of competing models that I fit separately to each participant's data. What is the correct way to calculate AIC in this case? Can I just write $AIC = 2nk-2\sum_{i=1}^n{LL_i}$, where $n$ ...
0
votes
0answers
21 views

AIC model selection in R. Transform and back-transform parameter estimates (slopes)

I have run an AIC model selection in R, (following Grueber etal 2011 J.Evo.Bio), and standardised my global model to a mean of 0 and SD of 2, using the "arm" package. AICc selection identified 2 ...
0
votes
0answers
26 views

AIC / BIC alternative in random effect model

I am looking at at panel data set. Hausmann recommends using random effects modeling, I have 3 nested models and Wald believes each step to add information. However, since some of the parameters had ...
1
vote
0answers
46 views

Using AIC or cross-validated MSE for selecting neural network models for time series prediction

I trained two basic feed-forward neural networks on time series data. The first one uses the observation at time step $t$ to predict $t+1$. Hence, it only has one predictor variable. The second ...
0
votes
0answers
33 views

Linear Discriminant Analysis for dimensionality reduction - choosing the dimension

I'm using Linear Discriminant Analysis to do dimensionality reduction of a multi-class data. What is the best method to determine the "correct" number of dimensions? Can I use a method similar to PCA, ...
3
votes
1answer
39 views

Model selection for random effects: can unselected random effects be used as fixed effects?

I am working on a mixed effects model. What I would consider random effects are year, sampling transect, and sampling location. There are multiple collections taken along each transect, and multiple ...
1
vote
1answer
28 views

Selecting model in GLM logistic with AICc, p-values for best AICc are all insignificant?

I have one potentially causal predictor and a number of covariates that I tested via AICc model selection in logistic GLM. I found that alone, the causal predictor has a low AICc (~19) and a ...
0
votes
0answers
15 views

Selecting best model in GLM logistic with AICc, p-values for best are insignificant?

I have one potentially causal predictor and a number of covariates that I tested via AICc model selection in logistic GLM. I found that alone, the causal predictor has a low AICc (~19) and a ...
1
vote
0answers
77 views

AIC comparison for models without maximum likelihood

AIC is a popular model comparison measure (despite of its potential shortcomings). I am wondering whether it is legit to compare AIC (or Akaike weights) for models that were fitted without requiring ...
1
vote
2answers
80 views

Interpretation of AIC value

Typical values of AIC that I have seen for logistic models are in thousands, at least hundreds. e.g. On http://www.r-bloggers.com/how-to-perform-a-logistic-regression-in-r/ the AIC is 727.39 While ...
0
votes
1answer
20 views

VAR lag selection heavily depends on maximum lag investigated

I am fitting an Error Correction Model with two monthly price time series. In Stata I am using the varsoc command to determine the number of lags that are ...
0
votes
0answers
17 views

Model selection for nested count data

What is the best tool-box for model selection when working with nested count data? Is AICc appropriate for comparing Poisson and negative binomial mixed models? Is there anything special to the ...
0
votes
1answer
47 views

Is multicollinearity an issue when doing stepwise logistic regression using AIC and BIC?

As far as I understood, it should not be a problem as long as I don't have perfect multicollinearity since I don't mind if the standard errors get inflated. However, what about using the ...
3
votes
2answers
73 views

Specify order of ARIMA model using autocorrelogram

How can I, using the correlograms above, specify the orders of the ARIMA model? These are the pac an ac of the differenced time series. Using AIC and BIC, I can't seem te find a proper model. ...
0
votes
0answers
19 views

Model comparison with different predictors

I have a conducted a series of experiments and manipulated a variable (X) that - from the literature - I know is relevant in this context. For theoretical reasons, I am now convinced that the ...
0
votes
0answers
42 views

finding best fit model using AIC in Panel Data using R

I am currently trying to find a best model using R for Panel Data. I have a project on Corporate Governance in which I collected data of various companies from 2009-2014. I found the best fit using ...
0
votes
2answers
38 views

Selecting best ARIMA model with regressors and dummy variable

I have data on GDP, employment rate, inflation and production on two countries and I like to make some ARIMA models. I have done this before, but not with including regressors. Also, the time period ...
0
votes
0answers
27 views

Generating SARIMA data and using it to evaluate the accuracy of the `predict` function, but getting some weird plots

I have written the following code to generate 500 data points from a $SARIMA$ model, use $400$ as training data and then predict the following $100$, while estimating the model with AIC. It appeared ...
2
votes
0answers
31 views

Generating an $ARMA(1,1)$ model with `arima.sim` in R, receiving warnings and errors but questioning the flaw in my method for estimation

So I wanted to generate $500$ data points from an $ARMA(1,1)$ distribution in R, use the first $400$ as my training data and use the training data and the predict ...
0
votes
0answers
26 views

Calculate AIC or Anova from an RJags model

I am fitting multivariate linear regressions with RJags (I have to do it with an mcmc because I'm taking all errors into account). I want to know between two polynomials which one fits better my ...
0
votes
0answers
50 views

Interpreting AIC forward stepwise function in R

My homework asks me to: "Try a forward stepwise procedure with entry probability of 0.20. Then describe the model that is arrived at and whether it might be preferred." I used the forward step ...
0
votes
0answers
10 views

Two “best models” (one is not-significant): Are they good models?

I am using GLM (poisson) to find temporal trends in my data. I have two "best" models (same AIC, same AICweight, same deviance) with one parameter each one. However one model is not significant. The ...
1
vote
1answer
27 views

How does one decide the most suitable GARCH model?

I am trying to model financial data using GARCH($p$,$q$). My question is, what information criteria do I use to determine which orders for $p$ and $q$ are most suitable? For instance, for ARIMA ...
3
votes
1answer
41 views

Match model selection strategies with modelling objectives

I am confused trying to match different model selection strategies with different modelling objectives. (Unfortunately, my confusion is reflected in the length of the post. Please be patient.) Model ...
0
votes
0answers
18 views

AIC of a mixture distribution

I'm trying to reproduce this paper: http://core.ac.uk/download/pdf/6394955.pdf where a latent class model/finite mixture model is used on the RAND Health insurance data. The data is freely available ...
6
votes
1answer
72 views

Is the Akaike information criterion inversely proportional to the chi-squared statistic?

I am in the process of calculating the akaike information criterion (AIC) for a set of 15 nested models. Data was generated from the 5th model and used in a parameter estimation for all models. This ...
0
votes
0answers
36 views

Zero-inflated Poisson regression Vuong test: Raw, AIC- or BIC-corrected results

I'm analyzing count data for a set of ten species and found that for the five species with highest detection rate, the zero-inflated poisson (ZIP) regression fits the data significantly better than ...
8
votes
2answers
178 views

AIC formula in Introduction to Statistical Learning

I'm a little puzzled by a formula presented in Hastie's "Introduction to Statistical Learning". In Chapter 6, page 212 (sixth printing, available here), it is stated that: $AIC = ...
0
votes
0answers
13 views

A robust version of AIC using S-estimators

Again a special question concerning a paper about robust versions of the AIC: Robust versions of AIC. I am trying to implement the AIC.S criterion described on page 7. This should be no problem ...
0
votes
0answers
22 views

AIC for S-estimation: A robust version of the AIC

I have got a very special question according an interesting paper about robust versions of the AIC. You find the paper here: Robust AIC In the paper there are a few modified robust criterions ...
0
votes
1answer
30 views

Criterion for model selection in an AR model?

I would like to smooth some financial time series data under the assumption that the data consists of variable trend and cyclic components plus white noise. I am thinking of applying an AR model to ...
0
votes
1answer
22 views

Parsimony vs AICc in ARIMA modelling

I'm trying to model daily financial data using an ARIMA model in R. After calculating returns, I used the auto.arima function and it chose an ARIMA(1,0,0) model as ...
0
votes
0answers
6 views

Do hypothesis-driven tests yield different conclusions depending on the order the models are computed?

whereas the AIC approach yields consistent results and is independent of the order in which the models are computed?
1
vote
0answers
64 views

AIC or BIC for robust regression?

I want to fit a robust regression method to my data because there are some outliers that might influence the estimates too much. Now my question: Are criterions like AIC or BIC still useful for robust ...
1
vote
1answer
47 views

Number of parameters in mixed model

How can I tell how many parameters will be estimated for random effects in mixed models? Here is the example from the lme function in nlme: ...
12
votes
2answers
435 views

Fundamental Question on Statistical Inference

I have been studying statistics from many books for the last 3 years, and thanks to this site I learned a lot. Nevertheless one fundamental question still remains unanswered for me. It may have a very ...
2
votes
0answers
149 views

Paradox in model selection (AIC, BIC, to explain or to predict?)

Having read Galit Shmueli's "To Explain or to Predict" (2010) I am puzzled by an apparent contradiction. There are three starting points, AIC- versus BIC-based model choice (end of p. 300 - start of ...
1
vote
1answer
57 views

Model selection between parametric nonparametric methods

I have a real data set (n=50). I would like to fit some parametric models to this data set and then compare the maximum log-likelihood values with my spline model which is a nonparametric model. could ...
0
votes
0answers
39 views

AIC in R: Back-transformation of model averaged coefficient estimates

I am running an analysis using a mixed model with lmer in R. I am using AIC as my model selection process. I have a global model and am including within the selection process all subset models ...
0
votes
0answers
68 views

AIC in R. How to back-transform model averaged coefficients?

I'm running an analysis in R for a mixed model (using lmer). ...
1
vote
1answer
54 views

AIC and anova p in multilevel model, how to interpret?

I have a model with a random-nested factor, I am comparing it with a model without the random factor (to test significance of random factor) as follows: ...
4
votes
1answer
94 views

Selecting regularization penalty: cross validation or information criteria?

I will use an elastic net to estimate a regression model which will later be used for forecasting. I have a grid of $\alpha$ values within [0,1] representing the proportion of $L_1$ versus $L_2$ ...
0
votes
0answers
26 views

Number of parameters of self-organizing map for AIC

What is the number of parameters of a self-organizing map (SOM) for calculating the AIC? Wikipedia cities Goutte (2001): "The k-means model is almost a Gaussian mixture model and one can construct a ...
0
votes
1answer
26 views

Question on practically using the Akaike information criterion

I am trying to develop the correct intuition about practically using the cost function AIC. I am under the impression that so long as the result of including an additional variable/parameter is a ...