Watched MIT 18.02 - lectures 9,10

The Second Derivative Test is a procedure for determining if a critical point is a maximum, a minimum, a saddle or a degenerate.

PROCEDURE:
- Calculate $f_x, f_y, f_{xx}, f_{xy}, f_{yy}$
- Calculate the critical points ;

Then for each critical point:
- Calculate $A = f_{xx}(x_0, y_0)$
- Calculate $B = f_{xy}(x_0, y_0)$
- Calculate $C = f_{yy}(x_0, y_0)$
- $AC-B^2$.

Apply the Second Derivative Test
If $AC-B^2 > 0$ and $A > 0$ minimum
If $AC-B^2 > 0$ and $A < 0$ maximum
If $AC-B^2 < 0$ saddle
If $AC-B^2 = 0$ degenerate

Click to enlarge

EXAMPLE:
( See Mathematica print )
$f(x,y)=e^{x^2-\frac{x^4}{4}-y^2}$

At $(0,0): \left\{AC-B^2, A\right\} = \left\{-4,2\right\}$
Saddle.

At $(-\sqrt{2},0): \left\{AC-B^2, A\right\} = \left\{8e^2,-4e\right\}$
Local maximum.

At $(\sqrt{2},0): \left\{AC-B^2, A\right\} = \left\{8e^2,-4e\right\}$
Local maximum.

.