Figure 2.

Schematic of linear classification. Each feature (x_i) of the data point (x₁,…,x_N) is multiplied by its respective weight (w_i), and the summation of the resulting terms ($\widehat{y}={\sum }_i{x}_i\cdot w_i$) is evaluated. The classifier predicts that the data point is in “class A” if $\widehat{y}\text{<}0$, and “class B” if $\widehat{y}>0$.