How to control for Nuisance variables in the logistic regression using fitglm(T,'​Distributi​on','binom​ial')?

How to control for Nuisance variables in the logistic regression using fitglm(T,'Distribution','binomial')?

Answer 1

[Vaibhav]

Hi Weixi

It is my understanding that you would like to include nuisance variables in the logistic regression using "fitglm" function.

Nuisance variables can be controlled by including them as additional predictor variables in the model. Including nuisance variables helps to account for their potential impact on the response variable and can improve the accuracy of the logistic regression model.

Here is an example to include nuisance variables in logistic regression using "fitglm":

% Generate some example data
rng(1); % For reproducibility
X1 = randn(100, 1);  % Predictor variable 1
X2 = randn(100, 1);  % Predictor variable 2
NuisanceVar = randn(100, 1);  % Nuisance variable
Y = randi([0, 1], 100, 1);  % Binary response variable

% Create a table with predictor and response variables
T = table(X1, X2, NuisanceVar, Y, 'VariableNames', {'X1', 'X2', 'NuisanceVar', 'Y'});

% Specify the logistic regression model
formula = 'Y ~ X1 + X2 + NuisanceVar';

% Fit the logistic regression model
mdl = fitglm(T, formula, 'Distribution', 'binomial');

% Display the model summary
disp(mdl);

Generalized linear regression model:
    logit(Y) ~ 1 + X1 + X2 + NuisanceVar
    Distribution = Binomial

Estimated Coefficients:
                   Estimate      SE        tStat     pValue 
                   ________    _______    _______    _______

    (Intercept)    0.18645     0.20492    0.90986     0.3629
    X1             0.12591     0.20649    0.60975    0.54203
    X2             0.11957     0.20778    0.57549    0.56496
    NuisanceVar    0.20014     0.23331    0.85781      0.391


100 observations, 96 error degrees of freedom
Dispersion: 1
Chi^2-statistic vs. constant model: 1.36, p-value = 0.716

You can refer to the following MathWorks documentation link to know more about "fitglm" function:

https://www.mathworks.com/help/stats/fitglm.html?s_tid=doc_ta

Hope this helps!