Maxima, Minima, and Saddle Points

Welcome to today’s lesson, students! 🎓 We’re diving into the fascinating world of maxima, minima, and saddle points for functions of two variables. By the end of this lesson, you’ll be able to:

Identify critical points of a function of two variables.
Classify those critical points as local maxima, local minima, or saddle points.
Apply the second derivative test for functions of two variables.
Connect these concepts to real-world situations, such as optimizing surfaces or economic models.

Ready to explore the peaks, valleys, and curious saddle-shaped points of multivariable calculus? Let’s get started! 🏔️🐎

Critical Points: The Starting Line

Before we can classify a point as a maximum, minimum, or saddle, we need to find it. These are called critical points.

Definition of Critical Points

A critical point of a function $f(x, y)$ is a point $(x_0, y_0)$ where both partial derivatives are zero (or where one or both partial derivatives do not exist). In other words, we find critical points by solving the system:

\frac{\partial f}{\partial x}(x_0, y_0) = 0 \quad \text{and} \quad \frac{\partial f}{\partial y}(x_0, y_0) = 0

These points are where the “slope” of the surface is flat in both the $x$ and $y$ directions—think of them as possible peaks, valleys, or flat spots.

Example: Finding Critical Points

Let’s consider the function:

f(x, y) = x^2 + y^2 - 4x - 6y + 13

We’ll find the critical points by taking the partial derivatives.

Compute the partial derivatives:

f_x = \frac{\partial f}{\partial x} = 2x - 4

f_y = \frac{\partial f}{\partial y} = 2y - 6

Set both partial derivatives to zero:

2x - 4 = 0 \quad \Rightarrow \quad x = 2

2y - 6 = 0 \quad \Rightarrow \quad y = 3

So, the critical point is at $(2, 3)$.

Great job! 🎯 Now we know where the critical point is. But is it a maximum, a minimum, or a saddle point? Let’s find out.

The Second Derivative Test: Classifying Critical Points

Once we have the critical points, we use the second derivative test to classify them. This test involves the second partial derivatives and something called the Hessian determinant.

The Hessian Matrix

The Hessian matrix $H$ for a function $f(x, y)$ is a 2x2 matrix of second partial derivatives:

$H = \begin{pmatrix}$

$f_{xx} & f_{xy} \\$

$f_{yx} & f_{yy}$

$\end{pmatrix}$

Where:

$f_{xx} = \frac{\partial^2 f}{\partial x^2}$ (the second partial derivative with respect to $x$)
$f_{yy} = \frac{\partial^2 f}{\partial y^2}$ (the second partial derivative with respect to $y$)
$f_{xy} = \frac{\partial^2 f}{\partial x \partial y} = f_{yx}$ (the mixed partial derivative)

The determinant of the Hessian matrix is called the discriminant $D$:

$D = f_{xx} f_{yy} - (f_{xy})^2$

The Second Derivative Test

At a critical point $(x_0, y_0)$:

If $D > 0$ and $f_{xx} > 0$, then $(x_0, y_0)$ is a local minimum.
If $D > 0$ and $f_{xx} < 0$, then $(x_0, y_0)$ is a local maximum.
If $D < 0$, then $(x_0, y_0)$ is a saddle point.
If $D = 0$, the test is inconclusive (we need other methods to classify the point).

Example: Classifying the Critical Point

Let’s go back to our example function:

f(x, y) = x^2 + y^2 - 4x - 6y + 13

We found a critical point at $(2, 3)$. Now let’s compute the second derivatives:

Compute the second derivatives:

f_{xx} = \frac{\partial^2 f}{\partial x^2} = 2

f_{yy} = \frac{\partial^2 f}{\partial y^2} = 2

f_{xy} = \frac{\partial^2 f}{\partial x \partial y} = 0

Calculate the Hessian determinant:

D = f_{xx} f_{yy} - (f_{xy})^2 = (2)(2) - (0)^2 = 4

We have $D = 4 > 0$ and $f_{xx} = 2 > 0$. According to the second derivative test, the critical point at $(2, 3)$ is a local minimum.

Intuition: Peaks, Valleys, and Saddles

Let’s visualize these concepts. Imagine you’re hiking ⛰️:

A local maximum is like the top of a hill—every direction you walk from that point goes downhill.
A local minimum is like the bottom of a valley—every direction you walk from that point goes uphill.
A saddle point is like a mountain pass—if you walk one way, you go uphill, but if you walk another way, you go downhill. It’s a mix of both.

Real-World Applications

Why do we care about maxima, minima, and saddle points? They show up everywhere in the real world!

Example 1: Economic Models

In economics, a company might want to maximize profit or minimize cost. A profit function $P(x, y)$ might depend on the amount of two products produced. Finding the maximum of $P(x, y)$ tells the company the optimal production levels to maximize profit.

Example 2: Engineering and Design

Engineers often need to find the shape of a surface that minimizes material use or maximizes strength. For example, the design of an airplane wing involves minimizing drag. This involves finding minima of complex functions that describe airflow.

Example 3: Machine Learning

In machine learning, algorithms often need to minimize a loss function. This is a function that measures how far off the model’s predictions are from the actual data. Finding the minima of the loss function helps create the best possible model.

Fun Fact: Saddle Points in Nature

Did you know that saddles aren’t just math concepts? 🐴 A real-world saddle point is often called a “col” in geography—like a mountain pass between two peaks. The famous Great St Bernard Pass in the Alps is a literal saddle point between Switzerland and Italy!

Another Example: A Saddle Point

Let’s look at a function that has a saddle point. Consider:

$f(x, y) = x^2 - y^2$

Find the partial derivatives:

f_x = 2x \quad \text{and} \quad f_y = -2y

Set them equal to zero:

2x = 0 \quad \Rightarrow \quad x = 0

-2y = 0 \quad \Rightarrow \quad y = 0

So, the critical point is at $(0, 0)$.

Find the second derivatives:

f_{xx} = 2, \quad f_{yy} = -2, \quad f_{xy} = 0

Calculate the Hessian determinant:

D = f_{xx} f_{yy} - (f_{xy})^2 = (2)(-2) - (0)^2 = -4

Since $D < 0$, the critical point at $(0, 0)$ is a saddle point.

Visualizing the Saddle Point

The graph of $f(x, y) = x^2 - y^2$ looks like a “pringle chip” or a horse saddle. If you walk along the $x$-axis, the function goes up in both directions (like a hill). But if you walk along the $y$-axis, the function goes down in both directions (like a valley). That’s the hallmark of a saddle point!

Multiple Critical Points

Sometimes functions have multiple critical points. Let’s look at a more complex example.

Example: Multiple Critical Points

Consider the function:

$f(x, y) = x^3 - 3xy^2$

Find the partial derivatives:

$ f_x = 3x^2 - 3y^2$

$ f_y = -6xy$

Set them equal to zero:

3x^2 - 3y^2 = 0 \quad \Rightarrow \quad x^2 = y^2 \quad \Rightarrow \quad y = $\pm$ x

-6xy = 0 \quad \Rightarrow \quad x = 0 \quad \text{or} \quad y = 0

So, we have two cases:

If $x = 0$, then $y = 0$. So, one critical point is $(0, 0)$.
If $x \neq 0$, then $y = \pm x$. So, we have two more critical points: $(1, 1)$ and $(-1, -1)$.

Classify the critical points using the second derivative test.

Let’s compute the second derivatives:

f_{xx} = 6x, \quad f_{yy} = -6x, \quad f_{xy} = -6y

At $(0, 0)$:

f_{xx}(0,0) = 0, \quad f_{yy}(0,0) = 0, \quad f_{xy}(0,0) = 0

The Hessian determinant is $D = (0)(0) - (0)^2 = 0$. The test is inconclusive here. We’ll need another method (like looking at the function’s graph) to classify this point.

At $(1, 1)$:

f_{xx}(1,1) = 6, \quad f_{yy}(1,1) = -6, \quad f_{xy}(1,1) = -6

D = (6)(-6) - (-6)^2 = -36 - 36 = -72

Since $D < 0$, $(1, 1)$ is a saddle point.

At $(-1, -1)$:

f_{xx}(-1,-1) = -6, \quad f_{yy}(-1,-1) = 6, \quad f_{xy}(-1,-1) = 6

D = (-6)(6) - (6)^2 = -36 - 36 = -72

Again, $D < 0$, so $(-1, -1)$ is also a saddle point.

Interpretation

This function $f(x, y) = x^3 - 3xy^2$ has three critical points: one at $(0, 0)$ (which we couldn’t classify with the second derivative test alone) and two saddle points at $(1, 1)$ and $(-1, -1)$. This shows us that functions can have multiple critical points with different behaviors.

Conclusion

Congratulations, students! You’ve mastered the essentials of finding and classifying maxima, minima, and saddle points for functions of two variables. 🏆

We started by finding critical points—those flat spots where both partial derivatives are zero. Then we used the second derivative test and the Hessian matrix to classify those points as local maxima, local minima, or saddle points. We saw real-world applications in economics, engineering, and machine learning. Finally, we explored examples with multiple critical points and encountered the fascinating geometry of saddle points.

Keep practicing, and soon you’ll be navigating the peaks and valleys of multivariable functions like a pro! 🚀

Study Notes

A critical point $(x_0, y_0)$ of $f(x, y)$ occurs where:

f_x(x_0, y_0) = 0 \quad \text{and} \quad f_y(x_0, y_0) = 0

The Hessian matrix $H$ is:

$ H = \begin{pmatrix}$

$ f_{xx} & f_{xy} \\$

$ f_{yx} & f_{yy}$

$ \end{pmatrix}$

The Hessian determinant (discriminant) is:

$ D = f_{xx} f_{yy} - (f_{xy})^2$

Second derivative test:

If $D > 0$ and $f_{xx} > 0$, local minimum.
If $D > 0$ and $f_{xx} < 0$, local maximum.
If $D < 0$, saddle point.
If $D = 0$, inconclusive.

Example: For $f(x, y) = x^2 + y^2 - 4x - 6y + 13$:
Critical point at $(2, 3)$.
$f_{xx} = 2$, $f_{yy} = 2$, $f_{xy} = 0$.
$D = 4 > 0$, and $f_{xx} > 0$, so $(2, 3)$ is a local minimum.

Example: For $f(x, y) = x^2 - y^2$:
Critical point at $(0, 0)$.
$f_{xx} = 2$, $f_{yy} = -2$, $f_{xy} = 0$.
$D = -4 < 0$, so $(0, 0)$ is a saddle point.

Real-world applications:
Maximize profit in economics.
Minimize material use in engineering.
Minimize loss functions in machine learning.

Keep these notes handy, students, and you’ll be ready to tackle any multivariable optimization problem that comes your way! 📚✨