Step 1 In simple linear regression method we find the relationship between two variables that is dependent variable and independent variable by using scatter diagram is the graphical method to check the relation between two variables. The simple linear regression equation is given by , \(Y = \beta_0 + \beta_1X\) Where Y is dependent variable X is inependent variable \(\beta_0\) is intercept of regression line \(\beta_1\) is the slope of the regression line In machine learning we called dependent variable (Y-variable) as target variable and independent (or Predictor)variable(X-variable) as feature vector.
Step 2 Yes ,we should omit a predictor from the modeling stage if it does not reflects any connection with the target variable in the EDA stage. Exploratory data analysis is the method of analyzing data sets to summarize their main characteristics within data visualization .This method was discovered by John Tukey. This step is important before you starting the machine learning or modelling of your data. In exploratory data analysis many graphical methods are available to check the relationship between two variables for example scatter plot, multi-vari chart, run chart ,pareto chart using the techniques we check the relationship between two variables if it does not show any relationship then we omit that predictor variable . Model specification is the method to determine which independent variables are included and excluded from the regression equation. Sometimes investigator measure too many variables but include some of them only and omit the variable that does not show any relationship with dependent or target variable .If investigator omits important variable from model the estimates for the variables that included can be biased and this is known as omitted variable bias . and it increase the bias in our model. To avoid bias in regression we omit variable that does not show any reflect connection with the target variable.