Читать книгу Fundamentals and Methods of Machine and Deep Learning - Pradeep Singh - Страница 18

Predicting values are more like solving an equation for the specified input. Consider an example where weight (y) is predicted from height (x). The LR equation is represented as [6]

(1.1)

or

(1.2)

These equations can be conspired as a line in 2-dimension as shown in Figure 1.2.

Figure 1.2 Height vs. weight graph [6].

Let B₀ be the bias coefficient and B₁ be the height column coefficient. To find the coefficients, the above learning techniques are used. Later, different height values are used to calculate the weight. For example, let B₀ = 0.2 and B₁ = 0.4, for an individual of height of 185 cm, the weight is calculated as follows [6]:

(1.3)

(1.4)