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
0answers
16 views

Information criteria for ARIMA model: missing log-likelihood for null model

I am trying to fit an ARIMA model on the time series of exchange rate. I have tried several kinds of ARIMA specifications (MA(1), MA(1,2), ...) and I am evaluating the particular setting according to ...
1
vote
1answer
30 views

Does AIC require the residuals of the model to be normally distributed?

Does AIC require the residuals of the models to be compared to be normally distributed?
1
vote
1answer
43 views

GLM model selection using AICc with Tweedie distribution

I have two questions regarding use of Tweedie GLM in R. I am new in using this distribution and despite a thorough search on different forums, I could not find my answers. I am now running several ...
0
votes
0answers
35 views

Comparing logistic model predictive ability with same sample but different IV

I would like to compare two logit models. Both models use the exact same sample (1 - pass a test and 0 - fail a test). The first model will use one IV (IV1) and the second model will use a different ...
0
votes
0answers
13 views

Sampling from a set of non-nested models

Consider a collection $\mathcal{M}$ of $m$ different model classes $\mathcal{M} = \{M_1,\dots,M_m\}$, where each model class has a parameter set $\Theta_i$, $i=1,\dots,m$. The model classes are not ...
2
votes
0answers
62 views

When AIC and Adjusted $R^2$ lead to different conclusions

I hope it's okay to ask theoretically driven R questions here. R has given me the following results from my 'tournament of models'. All models are entirely distinct except from 3 basic control ...
0
votes
1answer
29 views

Multiple linear regression: does BIC drop (vaguely) collinear variables?

Say I have the following multiple linear regression: Y ~ X1 + X2 + X3 + X4 All X variables are independent, but X1 and X2 look kind of linearly related when ...
0
votes
0answers
10 views

How to calculate Focused Information Criterion in R for Cox proportional hazards models? [migrated]

I am utilising R to perform a multivariate Cox survival regression for a research project. As I have many possible interchangeable variables in the model, I was wondering how to apply the Focused ...
1
vote
0answers
18 views

Estimating ARIMA models in SAS [migrated]

I would like to (insert "have to") do some ARIMA modeling in SAS. Normally, I would simply use auto.arima in R and let the function choose the differencing orders and just specify whether to use AICc ...
0
votes
0answers
23 views

How do I interpret which model is the best fit for my distribution (AIC)?

I'm new here so looking for some guidance; How do I interpret the following variables to understanding which model is considered the best fit to my distribution. ...
1
vote
0answers
14 views

Akaike information criterion for categorical and numerical data

How should I compute AIC for categorical and for numeric variables in classification problems? I see in Chapter 6 of Zumel and Mount that they use AIC before they train classification algorithms ...
1
vote
1answer
22 views

Alternative to AIC for feature selection in classification

I want to know what are the most common methods for feature selection in classification problems (binary and mutli-class). I see in Chapter 6 of Zumel and Mount that they use AIC before they train ...
0
votes
0answers
12 views

How to apply AIC to a situation where the mean of a multivariate normal is a 0-1 d-dimensional vector with exactly k 1's

I am trying to apply AIC to estimate mean in the following case: Let us consider that I have $n$ random variables $X_1, \ldots, X_n$, drawn i.i.d. from a normal distribution of mean $\mu\in\{0,1\}^d$ ...
1
vote
1answer
35 views

Model selection: can I compare the AIC from models of count data between linear and poisson models?

I am modeling count data (with offset / exposure parameter). My modeling strategy is use of a Poisson model and a negative binomial regression model. I compare model AICs, which are about -760 for my ...
0
votes
0answers
48 views

AIC versus cross validation in time series

I am interested in model selection in a time series setting. For concreteness, suppose I want to select an ARMA model from a pool of ARMA models with different lag orders. The ultimate intent is ...
2
votes
1answer
124 views

Selecting the number of mixtures / hidden states / latent variables

My question is regarding Gaussian Mixture models, Hidden Markov models (HMM) or any type of clustering or latent variable model, for which we can devise a likelihood function. Specifically, I train a ...
1
vote
1answer
185 views

Interpretation of (scale of) AIC, AICc and BIC when comparing different models

I'm trying to fit a model to a time series, but I am pretty confused as to which is the best. I'm looking at an arima model, and ets model and an stlf model, which each performed best within their ...
0
votes
1answer
54 views

Why would the residuals from these two models result in the wrong AIC being calculated?

We run two linear regressions, Model 1 and Model 2. The residuals from these two models are plotted against the predicted values. If I understand correctly, the AIC from these two models would be ...
0
votes
0answers
37 views

AIC/BIC values keeps falling as I add more and more lags. How do I select the appropriate lag length?

I am trying to minimize the values of the Akaike and Bayesian Information Criteria to figure out the optimal lag structure for my ARDL error correction model. I am using Stata to run my analysis and ...
0
votes
1answer
38 views

Parameter Estimation vs Inference Error

I am having trouble reconciling (or maybe even understanding properly) the following issues: We have a dataset. We hypothesize a functional form for probability density. Then we estimate the ...
1
vote
0answers
44 views

GLMM with 2 insignificant variables has lower AIC or BIC compared to same model without those variables…?

I am having a hard time understanding what's going on in with my model selection, and why a model with two insignificant variables is getting chosen as the "best model" over a model without those two ...
0
votes
1answer
35 views

comparing $R^2$ across two data sets

I have a set of covariates that characterize the type of experience a worker has (industry experience, general management experience, etc), and I am regressing compensation on these measures of ...
0
votes
0answers
111 views

Fixed term is “(Intercept)” --Dredge function

I'm currently building a glmer model for a resource selection analysis. See code below: ...
2
votes
1answer
131 views

Stepwise Model Selection in Logistic Regression in R

I'm implementing a logistic regression model in R and I have 80 variables to chose from. I need to automatize the process of variable selection of the model so I'm using the step function. I've no ...
2
votes
1answer
45 views

How is the Akaike Information Criteron applied for model with large number of predictors?

I am reading a paper (details not very relevant) which assumes that the market cost $C$ of a trade is related to $N$ predictors $X_1,\dots,X_N$ (page 25) through a linear relationship $$C = \beta_0 + ...
1
vote
1answer
53 views

On Negative AIC Values

My question is related to the thread Negative values for AIC in General Mixed Model. I often get negative AIC values from the software I use. I notice it most when I'm doing time series. But here ...
0
votes
1answer
73 views

AIC or BIC for model selection in Cox regression (SPSS)

Does anyone know if there is a method (a macro) to calculate AIC or BIC for a model in Cox regression using SPSS?
2
votes
1answer
84 views

The best model of an AICc-based model selection on a very small sample has an high number of predictors: does it make any sense?

I'm working with a very small sample size (N=14) and I'm using AICc to identify the most parsimonious model using a large number of possible predictors. Unexpectedly the best model has six predictors! ...
1
vote
1answer
103 views

how to decide which logistic regression model is better?

I have the following 3 models: ...
1
vote
1answer
110 views

lme4: Why is AIC no longer displayed when using REML [duplicate]

I have a simple question, understanding the basic usage of the lme4 package. I am following the tutorial by Bodo Winter ...
1
vote
1answer
47 views

Automatic selection of lowest information criterion comes with warning

I am building a forecasting model (ARMA) and found the very useful code-object arma_order_select_ic(see code below). It all works, however, each calculation comes ...
3
votes
2answers
261 views

What exactly is Box-Jenkins method for ARIMA process

The Wikipedia page says that Box Jenkins is a method of fitting an ARIMA model to a time series. Now, if I want to fit an ARIMA model to a time series, I will open up SAS, call proc ARIMA , supply the ...
0
votes
0answers
19 views

Is there much point in reducing the AIC penalty for a linear model to less than 2*p?

I'm currently using a Bayesian network, which, in this instance, is the same as a bunch of linear models. The sample size I have to work with is relatively low compared to the number of parameters, ...
0
votes
1answer
47 views

Comparing AIC or BIC for constant-only models vs ARIMA models

What if the AIC/BIC is lower (negatively speaking) with the model including just the constant with respect to other ARMA versions? I don't think because k=1 it is lower by construction.
0
votes
1answer
49 views

Is high AIC a bad feature of the model?

I have a model with AIC equal to 78 809. Does this mean this is a very bad model or the intepretation should be different? There are 15 variables, 2-level response variable and 40 000 rows. ...
0
votes
0answers
20 views

Generalized Estimating Equations- How many predictor variables are too many based on sample size?

I am conducting analyses on wild animals, on how diet of an individual changes based on environmental changes. I will list the setup of my dataset below. Until now, I have been running independent ...
1
vote
0answers
43 views

Definition of AIC in ARIMA() function in R?

I wonder how the Arima() function in R computes the AIC. Applying the standard formula AIC= 2*k - 2 LN(L) (with k number of parameters and L maximized value of likelihood) doesn't reproduce the ...
1
vote
1answer
58 views

A guide to regularization

I'm looking for some sort of guideline as when it is appropriate to use which forms of regularization and a comparison of the advantages / disadvantages of the various forms. So something that ...
0
votes
2answers
99 views

Structural equation modelling: model selection

I am currently trying to fit a structural equation model in R with the Lavaan package. I have this model that fits my data pretty good. This model is what I consider the full model, it has all paths ...
0
votes
0answers
72 views

AIC rankings: Why would a global model rank lower than an intercept-only model?

I'm working with some real-world (i.e. potentially messy) data on the nesting ecology of several bird species. I'm attempting to relate the daily survival rate of nests to vegetation characteristics ...
0
votes
1answer
98 views

Acceptable values for variance, aic and bic in multilevel models

I'm building a multilevel model from a sample of 820 observations at level 1 and 11 groups (level 2). I'm using stata xtmixed. Running the empty model (including ...
1
vote
0answers
30 views

Excluding Outliers and Influential Observations ($R^2$ and AIC/BIC)

I am working on a cross-sectional data set relating mortgage payments to debt-income ratios. I have some extreme outliers and experimented with excluding them from the model (some 30 observations of a ...
1
vote
0answers
59 views

Gamma vs tweedie distribution for large productivity dataset

I'm running some GAMs using the mgcv R package on a dataset with ~8.5k observations, where productivity is the response and environmental conditions are the covariates. However I am unsure of which ...
0
votes
2answers
67 views

Multiple linear regression, backward selection : Normality of the residuals?

I need to create a Multiple Linear regression model on those data explaining max03 T9 T12 T15 Ne9 Ne12 Ne15 Vx9 Vx12 Vx15 maxO3v !My data 1 My first intuition was to make a backward selection : ...
0
votes
0answers
33 views

“object 'logl.H0' not found” - Error in fitMeasures when calculating AIC for Lavaan model

EDIT: SOLVED The problem seems to have been an explanatory variable that was a factor. If it is made binary numeric insted, the values of BIC and AIC is calculated alright. However, the analyses give ...
0
votes
0answers
78 views

Akaike information criterion for Cox proportional hazard models

I am conducting an analysis of survival data using Cox proportional hazard (CPH) models, to figure out what is the best model to use. The models I am comparing are non-nested. My plan is to compute ...
0
votes
0answers
20 views

transforming of standardised effect size in MuMIN package

I ran the following model ...
6
votes
1answer
272 views

Should auto.arima in R ever report a model with higher AIC, AICC and BIC than other models considered?

I have used auto.arima to fit a time series model (a linear regression with ARIMA errors, as described on Rob Hyndman's site ) When finished - the output reports that the best model has a (5,1,0) ...
0
votes
0answers
18 views

AIC or similar selection techniques for Variograms?

I have a very basic question: how does one choose the "best" variogram? It is possible to fit different models to an empirical variogram, e.g. nugget, ...
5
votes
1answer
81 views

Assuming a probability density for MLE to do model selection

Motivation: I am trying to use Akaike Information Criterion to assess model ranking and over-fitting risk for a set of nonlinear models. I am an electrical engineer with no formal statistical training ...