Linear regression with larger data (rxFastLiner)

Linear regression is one of the most popular algorithms to predict the numeric outcome based on observed features. The default implementation in R for the linear regression is the lm() function. For a larger dataset with a large number of variables, this could take a very long time to run. The rxFastLinear() function for the RevoScaleR library offers a very fast implementation of linear regression with a larger dataset with many variables. In this recipe, you will build a linear regression model to predict arrival delay time as a function of the origin and destination airport along with the departure delay and the day of the week.

Getting ready

To build a linear regression model to predict the arrival delay time, you will need to have the RevoScaleR library. The dataset for this recipe will be the XDF file containing the following variables:

• YEAR
• QUARTER
• MONTH
• DAY_OF_MONTH
• DAY_OF_WEEK
• ORIGIN
• DEST
• DEP_DELAY
• ARR_DELAY

The objective is to build a linear...