Beyond Aggregation - Real-Time Analytics - page 462

Figure 11.2
Linear Models
Linear models are the most popular form of statistical modeling in many
fields, from economics to biology and beyond. (One of its first applications
was in the study of genetics.) This type of model is also known as a
regression model, the term used in this topic, as well as a least squares
model, which refers to the technique used to find the coefficients of the
model.
The essential idea is that an outcome variable, denoted as , has a linear
relationship with an array of explanatory variables, denoted as x[0] , x[1]
and so on, after the different elements of the array are multiplied by a
coefficient array, denoted as B[0] , B[1] , and so on:
public class LinearModel {
double [] B;
public LinearModel( double [] B) {
this .B = B;