Big Data and Business Analytics

Large- Scale Time- Series Forecasting • 201

amount of human intervention—highlighting the need for a generic fore-

casting methodology that is robust and provides an adequately accurate

forecast for each time series. In this section, we focus on how the google-

parallelism package in conjunction with Google’s infrastructure can be

a useful, practical, and inexpensive method for building, evaluating,

and engineering such a forecasting methodology in the R programming

language. A high- level overview of our forecasting methodology is pro-

vided in the next subsection, but further details are beyond the scope of

this chapter.

Brief Overview of Forecasting ...