I am currently trying to build one classifier using logistic regression algorithm.
mod<-step(glm(Dependent~., family = binomial(link=logit),data = train_final))
link - a specification for the model link function. This can be a name/expression, a literal character string, a length-one character vector or an object of class “link-glm” provided it is not specified via one of the standard names given next.
In this model, I want to know what does link specifies and how it’s value changes the model.
you want to do a classification and therefore you will have to use a sigmoid function to do a transformation. The glm delivers the log ratio not a classification (result of the linear regression) , the link function does the transformation from the log ratio to the binary class.
The link could change if you do not work with classification, for example if you want a ratio you will use a logit.
The wikipedia gives certainly a better explanation that i do check it enter link description here