Description
Consider a small dataset with the following points:
x y
1 2
2 3
3 5
4 4
We want to fit a linear model y = mx + b to this data using gradient descent.
1. Identify a good set of initial parameters for your model.
2. Compute the mean absolute error of your initial parameters:
E(m, b) = 1
n
Xn
i=1
|yi − (mxi + b)|
3. Make one adjustment to m and b using the derivative of E with respect to m and b.
What is your updated model?
4. compute the mean absolute error of your updated model.
