기록
Ordinary Least Squares Linear Regression: Flaws, Problems and Pitfalls 본문
machine learning
Ordinary Least Squares Linear Regression: Flaws, Problems and Pitfalls
민동기 2021. 10. 10. 14:22728x90
Problems and Pitfalls of Applying Least Squares Regression
1) outliers
2) non-linearities
3) too many variables
4) dependence among variables
5) wrong choice of error function
6) unequal training point variances (Heteroskedasticity)
7) wrong choise of features
8) noise in the independent variables
Ordinary Least Squares Linear Regression: Flaws, Problems and Pitfalls | An analysis of the defects of least squares regression.
LEAST squares linear regression (also known as “least squared errors regression”, “ordinary least squares”, “OLS”, or often just “least squares”), is one of the most basic and most commonly used prediction techniques known to humankind, wit
www.clockbackward.com
'machine learning' 카테고리의 다른 글
| K-SVD overview (0) | 2021.10.30 |
|---|---|
| ML lectures (0) | 2020.07.19 |
| sklearn에 있는 연습용 dataset (0) | 2020.07.19 |
| An overview of gradient descent optimization algorithms (0) | 2020.06.21 |
| Deep learning (0) | 2020.06.07 |