Regression in terms of Statistics vs. Computer Sciecne
|
Statistics |
Computer Science |
Method |
Algebra |
Update by Learning |
Parameter(Regression) |
OLS |
Delta Rule |
- Regression by Algebra(Statistics)
Goal: Find parameters that minimize OLS(Ordinary Least Squares)
- Regression by Learning(Computer Science)
Goal: Update parameters that minimize error (Y - Y’) given the epochs
When a = 0.1 and x, y are given, below table shows the procedure of updating parameter
data |
|
learning |
|
|
Update |
x |
y |
w |
y’ |
e |
delta_w=aex |
2 |
1 |
0.3 |
0.6 |
1 - 0.6 = 0.4 |
0.08 |
3 |
0 |
0.38 |
1.14 |
0- 1.14 = -1.14 |
-0.342 |
Multiple regression: MAE(Mean Absolute Error) and Paramters
- MAE changes as epoch increases