A decision function assigns a numerical score to a feature vector. For a linear classifier, the score combines feature values with weights and a bias; its sign determines the predicted class. section 8.2.1
The weights determine how features contribute to the score, while the bias shifts the decision boundary. section 8.2.1
Scoring two points
With f(x)=2x₁−x₂−3, the point (3, 1) scores 2 and receives class +1; (1, 4) scores −5 and receives class −1. The assignment of labels to real-world classes is arbitrary, provided it stays consistent. section 8.2.1
