A generalization of linear regression allowing for nonlinear relationships via a "link function" and for the variance of the response to depend on the predicted value. (Not to be confused with "general linear model" which extends the ordinary linear model to general covariance structure and ...

learn more… | top users | synonyms (1)

1
vote
2answers
11 views

credible interval equivalent of confint() for bayesglm() in Gelman et al's 'arm' package?

How do I extract a credible interval ala confint on a glm object when working with the object returned by bayesglm() in arm?
0
votes
1answer
10 views

R glmnet - appropriate link function for log-loss

I have a 2 class classification problem that I'm trying to optimize the log-loss for (not ROC/AUC). I'm using the glmnet package in ...
1
vote
0answers
11 views

Probabilistic forcasting

My question relates to probabilistic forecasting. How does one actually go about computing a forecast? Lets say I have some data that can be modelled by a specific distribution, and the values of the ...
0
votes
1answer
21 views

Is this a job for mixture of experts regression or semi-hidden markov models or something else?

Data I have several thousand timeseries each comprising around 365 data points. Browsing through a few of them, it looks like each timeseries consists of several regimes (different number f regimes ...
0
votes
1answer
48 views

How to use LASSO to select glm model gaussian

I have a small sample size n<20. I want to find which combination of 8 variables better predict y. I was using a stepAICc but it is suggested to away stepwise model selection. I have tried lars ...
0
votes
1answer
20 views

Effect of Lag time between explanatory variables and response on linear regression

I am curious as to how lag time between explanatory variables and response variables affect linear regression models. I am looking at some environmental data, mainly precipitation, temperatures, and ...
0
votes
1answer
29 views

is there a way to plot best glm model in model selection

I have run this glm model y~poly(xa,2)+poly(xb,2)+... Then have found the best fitting model using AICc. The best fitting model has a subset of the ...
4
votes
1answer
47 views

Feasibility of Negative Binomial Spatial Regression

I have a set of crime count data where it appears that the data take on a negative binomial distribution. I have had some success converting the dependent variable (a crime count) into a rate and then ...
0
votes
0answers
9 views

temporal pseudoreplication in testing possible combinations in logistic model [on hold]

I am interested in comparing mortality of animals when they are put in groups of a fixed size but of varying composition. Suppose I have 9 different genetic lines and I form the following categories: ...
0
votes
0answers
14 views

Linear regression with faster decrease in coefficient error/variance?

Suppose we have set of variables $Y$ and $X$, which know are related by a linear relation $y_i=\alpha x_i +\beta$, and important for us is to find $\alpha$ and $\beta$ and the error in estimating ...
5
votes
1answer
24 views

GLM for proportional data and underdispersion

I'd like asking your help to understand a statistical issue from my data set. I ran a GLM with proportional data, using a binomial distribution. However, I've found underdispersion in my model and I ...
0
votes
0answers
10 views

Negative Binomial GLM and temporal correlation

Data: I work with marine mammals stranding time series from 1976 to 2013. The idea was to model the number of marine mammals stranding on the beach (response variable) using year and month as an ...
2
votes
1answer
33 views

Mann Whitney test, unequal sample sizes, different shaped distributions

I was hoping someone could help, I am comparing proportional data in to two different groups. One group has a sample size of 22 and the other 530. The data are not normally distributed and have ...
2
votes
1answer
33 views

Generalized linear models and central limit theorem

If a comparison of treatment means can be made with ANOVA or GLM because it is assumed errors are normally distributed as suggested by the central limit theorem, why would it be necessary to implement ...
1
vote
0answers
20 views

Poisson model with fractions

I have a simple website with a home page that has 5 different images on it. All images have a fixed set of 'features' associated with them (size, color, position etc.,). When a visitor comes to the ...
0
votes
0answers
19 views

How to interpret $\beta$'s p-value in GLM? [duplicate]

I have a generalized linear model (GLM) with an identity link function. After fitting the data, I obtain elements in $\beta$ and their corresponding $p$-values from MATLAB's ...
1
vote
0answers
24 views

Confidence interval for psychometric binomial GLM model on more than one subject in R

I'm trying to estimate the confidence interval of a psychometric curve (binomial probit GLM), for a population (now only two subjects). Suppose I've subject "a" and subject "b", which performs ...
0
votes
0answers
5 views

Extracting random variable coefficients from model averaged objects

I'd like to extract the coefficients of the random effects from a model averaged object of class averaging created in the package ...
0
votes
1answer
30 views

optimal mean squared error in linear regression

How to determine whether a mean squared error is low or high? For example: In my linear regression problem if I get mean squared error as 21.67, how do I decide whether the error is low or high? Is ...
0
votes
0answers
10 views

Correlation between response and predictor due to variable sharing

Assume a process which acts on two entities, $a$ and $b$, to result in a further entity, $c$. It is considered that the amount of entity $a$ determines the amount of entity $c$ such that entity $a$ ...
1
vote
0answers
12 views

How to set the weights in GLM where the response is varaince

I have observations that are sample variances from a design with 2 factors, 2 levels in each. Although I do not observe the samples according to which these sample variances are computed, I do know ...
7
votes
2answers
261 views

What are some reasons iteratively reweighted least squares would not converge when used for logistic regression?

I've been using the glm.fit function in R to fit parameters to a logistic regression model. By default, glm.fit uses iteratively reweighted least squares to fit the parameters. What are some reasons ...
1
vote
0answers
23 views

In R: Gravity model; how to handle vectorized data from symmetrical matrix?

I have vectorized data from a symmetrical matrix with collaboration data. So the collaboration between country i and country j is the same as between country j and country i. I deleted the principal ...
7
votes
1answer
189 views

Fitting a heteroscedastic generalized linear model for binomial responses

I have data from the following experimental design: my observations are counts of the numbers of successes (K) out of corresponding number of trials (...
6
votes
0answers
114 views

Overdispersion and modeling alternatives in Poisson random effect models with offsets

I have run into a number of practical questions when modeling count data from experimental research using a within-subject experiment. I briefly describe the experiment, data, and what I have done so ...
2
votes
1answer
38 views

Overdispersion in poisson glm

When calculating the dispersion deviance/degrees freedom I get the value 1.8. Is it absolutly necessary to carry out the glm using quasipoisson? What is deemed 'significantly overdispersed' ?
0
votes
1answer
37 views

Marginal effects calculation in R: logit model [closed]

How do you calculate marginal effects of parameters of logit model in R uging package {glm}? Are following codes correct? ...
1
vote
2answers
55 views

How do I read this linear model output from R? [duplicate]

I normally use SPSS for my statistics, however after having some issues with violations I've had to try and run a linear model in R as apparently its more robust. Someone sent me the code that I ...
0
votes
0answers
14 views

mixed models/glm - What is the best way to evaulate if a variable in time can predict outcome?

I am interested in how the slopes of a variable measured at different time points (e.g. 0,1,2,3,4 months) can predict another outcome variable measured at 4 months. I have done this by creating ...
0
votes
0answers
6 views

How to compare factor levels in an averaged model

I'm running an analysis of my thesis data where I measured biodiversity, associated environmental variables and 2 categorical factors (region with 3 levels and garden type with 4 levels). I'm using ...
4
votes
1answer
62 views

Binomial glmm with a categorical variable with full successes

I am running a glmm with a binomial response variable and a categorical predictor. The random effect is given by the nested design used for the data collection. The data looks like this: ...
2
votes
0answers
13 views

How can I evaluate spatial autocorrelation in a binomial GLMM?

Following Dormann et al 2007 Ecography, I have employed a GLMM approach in R to account for spatial autocorrelation in a binomial regression model (logistic regression) that does not have random ...
0
votes
0answers
15 views

choosing the best structure of the random effects in a GLMM [duplicate]

I am trying to choose the best random effect structure in a GLMM, before starting with the fixed terms. To do that I include all the fixed effect and their interactions (beyond optimal model) and ...
3
votes
2answers
99 views

Predictor variables sum up to 1 but not necessarily correlated - is it a problem? [closed]

I am trying to fit hierarchical mixture model (using ML and MCMC, but this shouldn't matter) where the linear predictor part contains 17 independent variables. These are habitat variables: for each ...
1
vote
0answers
18 views

GLM and temporal correlation [closed]

I work with marine mammals stranding time series from 1976 to 2013. The idea was to model the number of marine mammals stranding on the beach (response variable) using year and month as an explanatory ...
2
votes
1answer
40 views

I need both quadratic and linear coefficients in a GLM with binary response. What's the best option? [closed]

I have three predictors and one response. What can I do if my response variable is binary?
-1
votes
0answers
21 views

Odds, log odds and odds ratios [duplicate]

I'm trying to get an understanding of what these three statistics are in the context of logistic regression: Odds Log odds Odds ratios Can anyone provide a short, intuitive summary of these three ...
1
vote
0answers
13 views

using speedglm with all variables and interactions in R

speedglm(y~.,data=df,family=binomial('logit')) won't work also trying: speedglm(y~(.)^7,data=df,family=binomial('logit')) for all interaction values. does not work. ...
2
votes
1answer
34 views

Gamma GLM predicting the second parameter of the Gamma

The gamma distribution has two parameters, I understand that the linear predictor predicts $\mu = g^{-1}(X\beta)$ where $g$ is the link function but how does the linear predictor specify the second ...
0
votes
0answers
7 views

What is the best way to test if an IV has a stronger effect on the DV than another IV that a subset of the first IV?

I wish to investigate the effect of general international experience of a corporation measured as the number of foreign subsidiaries it own vs. the effect of local experience in a region measured as ...
0
votes
0answers
14 views

What's the difference between error distribution and residual distribution in generalized linear models?

I have met with generalized linear model, but I'm confused with the errors and residuals? Can anyone help me out? I have got three questions. (1)what's the difference between error and residual? ...
5
votes
1answer
144 views

Definition of exponential family

I was reading these lecture notes and came accross the definition of GLM using exponential family of distribution. The latter seemed to have a bit of ambiguity, so I've checked that's it indeed the ...
3
votes
3answers
138 views

glm in R - which pvalue represents the goodness of fit of entire model?

I am running glms in R (generalised linear models). I thought I knew pvalues - until I saw that calling up a summary for a glm does not give you an overriding pvalue representative of the model as a ...
0
votes
1answer
57 views

glm in R - is the pvalue for Intercept important? Which pvalue represents the goodness of fit of entire model?

I am running glms in R (generalised linear models). I thought I knew pvalues - until I saw that calling up a summary for a glm does not give you an overriding pvalue representative of the model as a ...
3
votes
0answers
53 views

Fitting a glm to a zero inflated positive continuous response

I'm trying to fit R glm's to data sets where the response is zero inflated positive continuous. This is an example data set ...
1
vote
0answers
27 views

Using the confidence intervals to improve predictive model success rate

I'm trying to build a binomial predictive model based on glm. My overall prediction is very low, in the order of 60%. But when I go for the datapoints that have the both boundaries in one side, for ...
0
votes
0answers
37 views

Modeling continuous abundance data with a GLM in R: how to select the correct distribution family?

I have abundance data (counts) that I have standardized by area sampled, making them continuous. I would like to explain them with my two independent variables using a GLM but I am having trouble ...
0
votes
0answers
13 views

Poisson glm overdispersion. Calculating QAICc for model comparison

I'm having some difficulties with my glm . The models are overdispersed so I want to run them as quasipoisson. I would like to be able to compare the models using QAICc but the packages I have ...
0
votes
0answers
9 views

post hoc nested

Hy, need to do post hoc test for nested factor in SAS. I have two factors, A and B with B nested within A and I want to perform post hoc test in GLM in SAS. Is it possible? Stevan
0
votes
0answers
15 views

How to select GLM model and account for sampling time? [duplicate]

I'm trying to detect relationships between species abundances (counts) and time (years) for many species using either Negative Binomial or Poisson regressions (depending on degree of dispersion). ...