¿Por qué es ? (Una regresión lineal variable)

Nota: = Suma de cuadrados totales, = Suma de errores al cuadrado, y = Suma de cuadrados de la regresión. La ecuación en el título a menudo se escribe como:$SST$$SSE$$SSR$

$$\sum_{i=1}^n (y_i-\bar y)^2=\sum_{i=1}^n (y_i-\hat y_i)^2+\sum_{i=1}^n (\hat y_i-\bar y)^2$$

Pregunta bastante directa, pero estoy buscando una explicación intuitiva. Intuitivamente, me parece que tendría más sentido. Por ejemplo, supongamos que el punto tiene el valor y correspondiente y , donde es el punto correspondiente en la línea de regresión. También suponga que el valor medio de y para el conjunto de datos es . Entonces, para este punto particular i, , mientras que y . Obviamente, . ¿No se generalizaría este resultado a todo el conjunto de datos? No lo entiendo$SST\geq SSE+SSR$$x_i$y i = 3 y i ˉ y = 0 S S T = ( 5 - 0 ) 2 = 5 2 = 25 S S E = ( 5 - 3 ) 2 = 2 2 = 4 S $y_i=5$$\hat y_i=3$$\hat y_i$$\bar y=0$$SST=(5-0)^2=5^2=25$$SSE=(5-3)^2=2^2=4$$SSR=(3-0)^2=3^2=9$$9+4<25$

Sumar y restar da Entonces, tenemos que demostrar que∑ n i =

$$\begin{eqnarray*} \sum_{i=1}^n (y_i-\bar y)^2&=&\sum_{i=1}^n (y_i-\hat y_i+\hat y_i-\bar y)^2\\ &=&\sum_{i=1}^n (y_i-\hat y_i)^2+2\sum_{i=1}^n(y_i-\hat y_i)(\hat y_i-\bar y)+\sum_{i=1}^n(\hat y_i-\bar y)^2 \end{eqnarray*}$$ . Write n Σ i = 1 ( y i - y i ) ( y i - ˉ y ) = n Σ i = 1 ( y i - y i ) y i - ˉ $\sum_{i=1}^n(y_i-\hat y_i)(\hat y_i-\bar y)=0$ Por lo tanto, (a) los residuosaei=yi - y inecesidad de ser ortogonal a los valores ajustados,Σ n i = 1 (yi - y i) y i=0, y (b) la suma de las necesidades valores ajustados para que sea igual a la suma de la variable dependiente,Σ n i = 1 y $$\sum_{i=1}^n(y_i-\hat y_i)(\hat y_i-\bar y)=\sum_{i=1}^n(y_i-\hat y_i)\hat y_i-\bar y\sum_{i=1}^n(y_i-\hat y_i)$$ $e_i=y_i-\hat y_i$$\sum_{i=1}^n(y_i-\hat y_i)\hat y_i=0$ .$\sum_{i=1}^ny_i=\sum_{i=1}^n\hat y_i$

En realidad, creo (a) es más fácil de mostrar en notación matricial para la regresión múltiple general de que el caso única variable es un caso especial:

$$\begin{eqnarray*} e'X\hat\beta &=&(y-X\hat\beta)'X\hat\beta\\ &=&(y-X(X'X)^{-1}X'y)'X\hat\beta\\ &=&y'(X-X(X'X)^{-1}X'X)\hat\beta\\ &=&y'(X-X)\hat\beta=0 \end{eqnarray*}$$ As for (b), the derivative of the OLS criterion function with respect to the constant (so you need one in the regression for this to be true!), aka the normal equation, is $$\frac{\partial SSR}{\partial\hat\alpha}=-2\sum_i(y_i-\hat\alpha-\hat\beta x_i)=0,$$ $$\sum_i y_i=n\hat\alpha+\hat\beta\sum_ix_i$$ The right hand side of this equation evidently also is $\sum_{i=1}^n\hat y_i$, as $\hat y_i=\hat\alpha+\hat\beta x_i$.

(1) Intuition for why $SST = SSR + SSE$

When we try to explain the total variation in Y ($SST$) with one explanatory variable, X, then there are exactly two sources of variability. First, there is the variability captured by X (Sum Square Regression), and second, there is the variability not captured by X (Sum Square Error). Hence, $SST = SSR + SSE$ (exact equality).

(2) Geometric intuition

Please see the first few pictures here (especially the third): https://sites.google.com/site/modernprogramevaluation/variance-and-bias

Some of the total variation in the data (distance from datapoint to $\bar{Y}$) is captured by the regression line (the distance from the regression line to $\bar{Y}$) and error (distance from the point to the regression line). There's not room left for $SST$ to be greater than $SSE + SSR$.

(3) The problem with your illustration

You can't look at SSE and SSR in a pointwise fashion. For a particular point, the residual may be large, so that there is more error than explanatory power from X. However, for other points, the residual will be small, so that the regression line explains a lot of the variability. They will balance out and ultimately $SST = SSR + SSE$. Of course this is not rigorous, but you can find proofs like the above.

Also notice that regression will not be defined for one point: $b_1 = \frac{\sum(X_i -\bar{X})(Y_i-\bar{Y}) }{\sum (X_i -\bar{X})^2}$, and you can see that the denominator will be zero, making estimation undefined.

Hope this helps.

--Ryan M.

When an intercept is included in linear regression(sum of residuals is zero), $SST=SSE+SSR$.

prove

$$\begin{eqnarray*} SST&=&\sum_{i=1}^n (y_i-\bar y)^2\\&=&\sum_{i=1}^n (y_i-\hat y_i+\hat y_i-\bar y)^2\\&=&\sum_{i=1}^n (y_i-\hat y_i)^2+2\sum_{i=1}^n(y_i-\hat y_i)(\hat y_i-\bar y)+\sum_{i=1}^n(\hat y_i-\bar y)^2\\&=&SSE+SSR+2\sum_{i=1}^n(y_i-\hat y_i)(\hat y_i-\bar y) \end{eqnarray*}$$ Just need to prove last part is equal to 0: $$\begin{eqnarray*} \sum_{i=1}^n(y_i-\hat y_i)(\hat y_i-\bar y)&=&\sum_{i=1}^n(y_i-\beta_0-\beta_1x_i)(\beta_0+\beta_1x_i-\bar y)\\&=&(\beta_0-\bar y)\sum_{i=1}^n(y_i-\beta_0-\beta_1x_i)+\beta_1\sum_{i=1}^n(y_i-\beta_0-\beta_1x_i)x_i \end{eqnarray*}$$ In Least squares regression, the sum of the squares of the errors is minimized. $$SSE=\displaystyle\sum\limits_{i=1}^n \left(e_i \right)^2= \sum_{i=1}^n\left(y_i - \hat{y_i} \right)^2= \sum_{i=1}^n\left(y_i -\beta_0- \beta_1x_i\right)^2$$ Take the partial derivative of SSE with respect to $\beta_0$ and setting it to zero. $$\frac{\partial{SSE}}{\partial{\beta_0}} = \sum_{i=1}^n 2\left(y_i - \beta_0 - \beta_1x_i\right)^1 = 0$$ So $$\sum_{i=1}^n \left(y_i - \beta_0 - \beta_1x_i\right)^1 = 0$$ Take the partial derivative of SSE with respect to $\beta_1$ and setting it to zero. $$\frac{\partial{SSE}}{\partial{\beta_1}} = \sum_{i=1}^n 2\left(y_i - \beta_0 - \beta_1x_i\right)^1 x_i = 0$$ So $$\sum_{i=1}^n \left(y_i - \beta_0 - \beta_1x_i\right)^1 x_i = 0$$ Hence, $$\sum_{i=1}^n(y_i-\hat y_i)(\hat y_i-\bar y)=(\beta_0-\bar y)\sum_{i=1}^n(y_i-\beta_0-\beta_1x_i)+\beta_1\sum_{i=1}^n(y_i-\beta_0-\beta_1x_i)x_i=0$$ $$SST=SSE+SSR+2\sum_{i=1}^n(y_i-\hat y_i)(\hat y_i-\bar y)=SSE+SSR$$

This is just the Pythagorean theorem!