Machine Learning for Developers

Formalizing our concepts

Now let's go over the mathematical side of our process, so we have a reference of all the parts involved, before using them in practice.

The elements or our equations are as follows:

• The linear function variables, β₀ and β₁ 
• The number of samples in the sampleset, m
• The different elements in the sampleset, x⁽ⁱ⁾ and y⁽ⁱ⁾

Let's start with our error function, J. It was defined in previous sections under the name of the least squares function. We will add, for practicality, the term 1/2m at the start of the equation, as follows:

Least squares error function

Let's introduce a new operator, which is the basis of all the ...