Generalized method of moments: 두 판 사이의 차이

편집 요약 없음
편집 요약 없음
 
5번째 줄: 5번째 줄:
[[Ordinary least square (OLS)]] and [[instrumental variable (IV)]] estimators can be seen as special cases of GMM. In an OLS regression <math>y_i = \beta x_i + \epsilon_i</math>, the assumption <math>\mathbb{E}(x_i \epsilon_i) = 0</math> is a moment condition. The OLS estimator can be derived by setting the sample analogue:
[[Ordinary least square (OLS)]] and [[instrumental variable (IV)]] estimators can be seen as special cases of GMM. In an OLS regression <math>y_i = \beta x_i + \epsilon_i</math>, the assumption <math>\mathbb{E}(x_i \epsilon_i) = 0</math> is a moment condition. The OLS estimator can be derived by setting the sample analogue:


<div style="text-align: center;">
<math>\frac{1}{N}\sum_i x_i (y_i - \hat\beta x_i) = 0</math>
<math>\frac{1}{N}\sum_i x_i (y_i - \hat\beta x_i) = 0</math>
</div>


== References ==
== References ==
<references />
<references />