Regression with lines and parabolas
- Ingolf Terveer
This applet illustrates the construction of an optimal regression function for linear and quadratic regression.
Try to fit the function type to the given data. For linear functions equivalently, match the intercept (the mean point M must be on the line). For quadratic functions, the intercept and the vertex V has to be adjusted. To help adjustment, fade in the Sum of squares and observe its trace when changing intercept resp. vertex. If the point P is outside the plotted region, you have to use the scale bars. Finally, you can check on your guesses by showing the optimal solution