Lecture 6 Flashcards
1
Q
What is the derivative of R_sq(w_0, w_1)?
A
-2/n En i=1 (y_i - (w_0 + w_1x_i))
2
Q
What is w_0*?
A
ybar - w_1* xbar
3
Q
What is w_1*?
A
(En i =1 (y_i - ybar)x_i) / (En i=1 (x_i - xbar)x_i)
4
Q
What is the equation for xbar?
A
1/n En i=1 x_i
5
Q
What is the equation for ybar?
A
1/n En i=1 y_i
6
Q
What are the least squares solution for w_1*?
A
(En i=1 (x_i - xbar)(y_i - ybar)) / (En i=1(x_i - xbar)^2)
7
Q
What are the least squares solution for w_0*?
A
ybar - w_1*xbar
8
Q
What do we call w_0* and w_1*? the optimal parameters
A
optimal parameters
9
Q
What do we call the resulting line?
A
the regression line
10
Q
What does “fitting to the data” mean?
A
the process of minimizing empirical risk to find the optimal parameters
11
Q
What equation do we use to make predictions about the future?
A
H(x) = w_0 + w_1*x
