91Ó°ÊÓ

Partial derivatives are like regular derivatives, but they apply to functions with more than one variable. For a function of two variables, say \(f(x, y)\), the partial derivative with respect to \(x\) (denoted as \(f_x\)) is found by holding \(y\) constant and differentiating with respect to \(x\). Similarly, the partial derivative with respect to \(y\) (denoted as \(f_y\)) is found by holding \(x\) constant and differentiating with respect to \(y\).
Such derivatives are useful in determining how the function changes at a given point along each variable’s direction. In the context of critical points, partial derivatives are used to identify where the slope of the surface is flat, which could potentially indicate a local maximum, minimum, or a saddle point.
At a critical point, both \(f_x\) and \(f_y\) are equal to zero, suggesting no local slope and setting the stage for further investigation with the second derivative test.

The second derivative test helps us to determine the nature of a critical point by examining the concavity of the function. This involves using second-order partial derivatives: \(f_{xx}\), \(f_{yy}\), and \(f_{xy}\).
The second derivative \(f_{xx}\) is the partial derivative of \(f_x\) with respect to \(x\), \(f_{yy}\) is the partial derivative of \(f_y\) with respect to \(y\), and \(f_{xy}\) or \(f_{yx}\) is the mixed partial derivative, which in most cases are equal due to Clairaut's theorem.

• If both \(f_{xx}\) and \(f_{yy}\) are positive, the function is concave up at the critical point, hinting at a local minimum.
• If both are negative, the function is concave down, indicating a local maximum.

The test involves computing these second partial derivatives at the critical point to further proceed with understanding its nature.

The discriminant in calculus involves using a specific formula that incorporates the second partial derivatives to classify critical points. The formula for the discriminant \(D\) is:
\[ D = f_{xx}(a, b) \cdot f_{yy}(a, b) - [f_{xy}(a, b)]^2 \]
This discriminant helps in determining the nature of the critical point \((a, b)\):

• If \(D > 0\) and \(f_{xx} > 0\), we have a local minimum.
• If \(D > 0\) and \(f_{xx} < 0\), we have a local maximum.
• If \(D < 0\), the point is a saddle point.
• If \(D = 0\), the test is inconclusive, requiring further investigation.

This process is crucial for determining the type of extremum or feature the function has at the critical point.

Saddle points occur at critical points where \(D < 0\), suggesting neither a local maximum nor a minimum. The term comes from the way a horse saddle curves up in one direction and down in another. Saddle points often indicate a change in the nature of a function at a critical point.
They illustrate cases where the function curves upwards in one direction, and simultaneously curves downwards in an orthogonal direction. This characteristic ensures the point is genuinely flat (as both first partial derivatives are zero), yet it is not an extremum.
Recognizing a saddle point is crucial in multivariable calculus and critical in optimization problems, as it can often be misleading if one only looks at the fact that the partial derivatives are zero. Therefore, understanding additional factors, such as the discriminant, is important in distinguishing these critical points.