Noise is unwanted data items, features or records which don$$t help in explaining the feature itself, or the relationship between feature & target. Noise often causes the algorithms to miss out patterns in the data.

suppose $,$we want to perform a regression analysis with multiple variables.in some cases When we are using all variables in building a model our accuracy percentage will be less.which means some of the variables are not related to the dependent variable or no relationship with it$.$so simply we can term it as noise variables.when we remove these variables,we can get model with highly related dependent variable.