Linear Algebra - Classifying Quadratic Forms

1. Formula
   Observation

   The handwritten examples repeatedly use Q to denote a quadratic form, such as $Q = x_1^2$ and $Q = x_1^2 + x_2^2$.

2. Audio
   Observation

   The speaker says, "if I have the quadratic form Q equals x one squared" and later refers to "a quadratic form Q of x."

Q

A quadratic form; in the worked examples it is written as an explicit polynomial in variables $x_1$ and $x_2$.

In this clip, Q is shown for one-variable and two-variable quadratic forms.

1. Formula
   Observation

   $x_1$ appears in all four handwritten examples: $Q = x_1^2$, $Q = -2x_1^2$, $Q = x_1^2 + x_2^2$, and $Q = x_1^2 - x_2^2$.

2. Audio
   Observation

   The speaker names the variable as "x one" when reading the examples.

$x_1$

First coordinate variable of the quadratic form input.

1. Formula
   Observation

   $x_2$ appears in the two-variable examples $Q = x_1^2 + x_2^2$ and $Q = x_1^2 - x_2^2$.

2. Audio
   Observation

   The speaker says, "say Q equals x one squared plus x two squared" and later "Q equals x one squared minus x two squared."

$x_2$

Second coordinate variable of the quadratic form input.

1. Audio
   Observation

   The speaker says, "think about this as a function f of x equals x squared."

1. The formula $f(x)=x^2$ is spoken but not clearly written on screen.

$f(x)$

An ordinary single-variable function used to reinterpret the one-variable quadratic form for graphing.

1. Formula
   Observation

   The definition block writes "A quadratic form $Q(x)$ is:" and then uses $Q(x)$ in each bullet condition.

2. Audio
   Observation

   The speaker says, "We say that a quadratic form Q of x is positive definite..."

$Q(x)$

The value of the quadratic form evaluated at an input vector x.

1. Formula
   Observation

   The definition bullets use x with a vector arrow, including the condition $x \neq \vec{0}$.

2. Audio
   Observation

   The speaker says, "whenever x is not the zero vector."

$\vec{x}$

Input vector to the quadratic form.

1. Formula
   Observation

   The definition bullets explicitly exclude $\vec{x}=\vec{0}$ by writing $\vec{x}\neq\vec{0}$.

2. Audio
   Observation

   The speaker says, "whenever x is not the zero vector" and "for all x's that's not the zero vector."

$\vec{0}$

The zero vector, which is excluded from the definiteness conditions.

1. Formula
   Observation

   The board repeatedly writes $Q(x)$ in the definition and theorem statements.

$Q(\vec{x})$

A quadratic form evaluated at vector $\vec{x}$.

$\vec{x}$ is a vector; the video does not explicitly state the ambient dimension.

1. Formula
   Observation

   The definition uses 'for all $\vec{x} \neq \vec{0}$'.

$\vec{0}$

The zero vector, excluded in the strict definiteness conditions.

Vector space containing $\vec{x}$; exact dimension not stated.

1. Formula
   Observation

   The theorem line reads 'If $Q = \vec{x}^T A \vec{x}$, then ...' and the audio says 'where A is a symmetric matrix'.

A

The symmetric matrix associated with the quadratic form Q.

Square matrix; size matches the dimension of $\vec{x}$.

1. Formula
   Observation

   The displayed theorem condition is $Q = \vec{x}^T A \vec{x}$.

$\vec{x}^T$

Transpose of the vector $\vec{x}$, used to form the scalar quadratic expression.

Row-vector form of $\vec{x}$.

1. Formula
   Observation

   The theorem bullets refer to 'the eigenvalues of A'.

2. Audio
   Observation

   The lecturer says classification is done by finding the eigenvalues of the symmetric matrix A.

$\text{eigenvalues of }A$

Scalars associated with the symmetric matrix A that determine the sign type of Q.

Real eigenvalues are implied by symmetry, but the video does not separately prove this fact.

1. Audio
   Observation

   The speaker says, "Now one way to classify these quadratic forms is by the signs of the output when you plug in various points for x one and x two."

The clip introduces a classification method based on whether the values produced by a quadratic form are positive, negative, or both when different nonzero inputs are substituted.

1. The method is presented for quadratic forms in variables such as $x_1$ and $x_2$.

2. It is introduced after several graphed examples.

1. Formula
   Observation

   The definition block writes: "positive definite if $Q(\vec{x}) > 0$ for all $\vec{x} \neq \vec{0}$."

2. Audio
   Observation

   The speaker says, "We say that a quadratic form Q of x is positive definite if Q of x is greater than zero whenever x is not the zero vector."

A quadratic form is called positive definite when its value is strictly greater than zero for every nonzero input vector.

1. The inequality must hold for all $\vec{x} \neq \vec{0}$.

2. The zero vector is excluded from the condition.

1. Classifying quadratic forms by output signs

1. Formula
   Observation

   The definition block writes: "negative definite if $Q(\vec{x}) < 0$ for all $\vec{x} \neq \vec{0}$."

2. Audio
   Observation

   The speaker says, "We say that a quadratic form Q of x is negative definite if Q of x is less than zero for all x's that's not the zero vector."

A quadratic form is called negative definite when its value is strictly less than zero for every nonzero input vector.

1. The inequality must hold for all $\vec{x} \neq \vec{0}$.

2. The zero vector is excluded from the condition.

1. Classifying quadratic forms by output signs

1. Formula
   Observation

   The definition block writes: "indefinite if $Q(\vec{x})$ has both positive and negative values."

2. Audio
   Observation

   The speaker says, "Finally, we say that the quadratic form Q of x is indefinite if Q of x has both positive and negative values."

A quadratic form is called indefinite when it takes both positive and negative values among its outputs.

1. There must exist inputs producing positive values and other inputs producing negative values.

2. The clip does not state a separate exclusion clause for $\vec{x}=\vec{0}$ in this bullet.

1. Classifying quadratic forms by output signs

1. Audio
   Observation

   The lecturer says a quadratic form is positive semidefinite if $Q(x)$ is greater than or equal to zero.

2. Formula
   Observation

   Blue text appears: 'positive semidefinite if $Q(\vec{x}) \ge 0$.'

A quadratic form $Q(\vec{x})$ is called positive semidefinite when its value is never negative: $Q(\vec{x}) \ge 0$ for every input vector. Unlike positive definite, equality at nonzero vectors is allowed.

1. Applies to a quadratic form $Q(\vec{x})$.

2. No explicit exclusion of $\vec{x}=\vec{0}$ is stated in the semidefinite line.

1. Quadratic form sign classification setup

1. Audio
   Observation

   The lecturer says a quadratic form is negative semidefinite if $Q(x)$ is less than or equal to zero.

2. Formula
   Observation

   Blue text appears: 'negative semidefinite if $Q(\vec{x}) \le 0$.'

A quadratic form $Q(\vec{x})$ is called negative semidefinite when its value is never positive: $Q(\vec{x}) \le 0$ for every input vector. This is the sign-reversed analogue of positive semidefinite.

1. Applies to a quadratic form $Q(\vec{x})$.

2. No explicit exclusion of $\vec{x}=\vec{0}$ is stated in the semidefinite line.

1. Quadratic form sign classification setup

1. Audio
   Observation

   The lecturer says we can classify these quadratic forms by writing Q in the form x transpose times A times x and finding the eigenvalues of the symmetric matrix A.

2. Formula
   Observation

   At 30s the theorem line begins with 'If $Q = \vec{x}^T A \vec{x}$, then'.

To classify a quadratic form, rewrite it as $Q=\vec{x}^T A\vec{x}$ with A symmetric, then inspect the signs of the eigenvalues of A. The sign pattern of those eigenvalues determines whether Q is positive definite, negative definite, or indefinite.

1. A must be symmetric.

2. The method applies to quadratic forms that can be represented in this matrix form.

1. Quadratic form sign classification setup
2. Positive definite quadratic form
3. Negative definite quadratic form
4. Indefinite quadratic form

1. Formula
   Observation

   The black definition block states: 'A quadratic form $Q(\vec{x})$ is:' followed by three bullet cases.

1. The clip does not give a separate algebraic definition of what makes an expression a quadratic form; it only defines sign classes of $Q(\vec{x})$.

The board treats $Q(\vec{x})$ as a quadratic form and classifies it by the signs of its values over nonzero vectors. The three strict classes shown first are positive definite, negative definite, and indefinite.

1. The strict definitions quantify over all $\vec{x} \neq \vec{0}$.

1. Formula
   Observation

   Black text: 'positive definite if $Q(\vec{x}) > 0$ for all $\vec{x} \neq \vec{0}$.'

A quadratic form $Q(\vec{x})$ is positive definite if it takes strictly positive values on every nonzero input vector.

1. Quantified over all nonzero vectors $\vec{x}$.

1. Quadratic form sign classification setup

1. Formula
   Observation

   Black text: 'negative definite if $Q(\vec{x}) < 0$ for all $\vec{x} \neq \vec{0}$.'

A quadratic form $Q(\vec{x})$ is negative definite if it takes strictly negative values on every nonzero input vector.

1. Quantified over all nonzero vectors $\vec{x}$.

1. Quadratic form sign classification setup

1. Formula
   Observation

   Black text: 'indefinite if $Q(\vec{x})$ has both positive and negative values.'

A quadratic form $Q(\vec{x})$ is indefinite if it assumes both positive and negative values as $\vec{x}$ varies.

1. No explicit nonzero-vector qualifier is written in this bullet, unlike the definite cases.

1. Quadratic form sign classification setup

1. Formula
   Observation

   Definition: A quadratic form $Q(x)$ is: positive definite if $Q(x)>0$ for all x!=0; negative definite if $Q(x)<0$ for all x!=0; indefinite if $Q(x)$ has both positive and negative values.

The video defines three classifications for a quadratic form $Q(x)$: positive definite ($Q(x) > 0$ for all non-zero x), negative definite ($Q(x) < 0$ for all non-zero x), and indefinite ($Q(x)$ takes both positive and negative values). It also mentions positive semi-definite and negative semi-definite.

1. Applies to quadratic forms $Q(x)$

1. Formula
   Observation

   The theorem block reads: 'Theorem: If $Q = \vec{x}^T A \vec{x}$, then' followed by three iff statements about eigenvalues of A.

2. Audio
   Observation

   The lecturer states each equivalence in words: positive definite iff all eigenvalues positive, negative definite iff all eigenvalues negative, indefinite iff both positive and negative eigenvalues.

If $Q=\vec{x}^T A\vec{x}$ with A symmetric, then Q is positive definite iff all eigenvalues of A are positive; Q is negative definite iff all eigenvalues of A are negative; and Q is indefinite iff A has both positive and negative eigenvalues.

1. Q can be written as $\vec{x}^T A\vec{x}$.

2. A is a symmetric matrix.

Universal over the eigenvalues in the definite cases; existential mixture in the indefinite case ('both positive and negative eigenvalues').

1. Audio
   Observation

   The lecturer says the proof boils down to making a change of variables so Q can be rewritten without cross-product terms.

2. Audio
   Observation

   He then says Q becomes lambda_1 $y_1^2$ + lambda_2 $y_2^2$ + ... + lambda_n $y_n^2$, and uses the signs of the lambda_i to explain the theorem.

1. The full proof is explicitly omitted in the video.

2. The diagonalized formula is spoken only, not written on the board.

The theorem is proved by a change of variables that rewrites Q as a sum of pure squares with coefficients equal to the eigenvalues of A, after which the sign of Q follows from the signs of those coefficients.

1. A is symmetric.

2. $Q=\vec{x}^T A\vec{x}$.

For an n-variable quadratic form, the diagonalized expression runs over $i=1$,$\dots$,n.

1. Audio
   Observation

   The speaker says, "First off, we can graph quadratic forms by thinking of them as functions."

2. Animation
   Observation

   The board successively shows a parabola, an inverted parabola, a paraboloid, and a hyperbolic saddle surface corresponding to the four written examples.

1. Expression

   $$Q = x_1^2$$
   Explanation

   The first example is treated as a one-variable quadratic form.

   Justification

   This is directly written on the board and read aloud.

   Shown in the video
2. Expression

   $$f(x)=x^2$$
   Explanation

   The speaker reinterprets the one-variable form as an ordinary function so it can be graphed in the plane.

   Justification

   Stated verbally in the audio.

   Shown in the video
3. Expression

   $$Q = -2x_1^2$$
   Explanation

   Changing the coefficient to a negative number produces an inverted parabola.

   Justification

   The speaker says this "is going to look like an inverted parabola," and the drawing matches.

   Shown in the video
4. Expression

   $$Q = x_1^2 + x_2^2$$
   Explanation

   With two variables, the graph is drawn in three dimensions as a paraboloid.

   Justification

   The speaker says, "then I can draw this in three dimensions, and the shape that we have is a paraboloid."

   Shown in the video
5. Expression

   $$Q = x_1^2 - x_2^2$$
   Explanation

   Replacing the plus sign with a minus sign changes the surface into a hyperbolic surface with a saddle point.

   Justification

   The speaker says, "then we have a hyperbolic surface" and adds that such surfaces have saddle points in the middle.

   Shown in the video

The clip uses graphing as an intuitive bridge from algebraic formulas to geometric shapes before introducing formal sign-based classifications.

1. Audio
   Observation

   The lecturer explains that after a change of variables, Q looks like lambda_1 $y_1^2$ + lambda_2 $y_2^2$ + ... + lambda_n $y_n^2$.

2. Audio
   Observation

   He then reasons case by case about all-positive eigenvalues, all-negative eigenvalues, and mixed signs.

1. This is a proof sketch, not a full rigorous derivation.

2. The diagonalized expression is not shown visually.

1. Expression

   $$Q=\vec{x}^T A\vec{x}$$
   Explanation

   Start from the matrix representation of the quadratic form with symmetric A.

   Justification

   Stated in the theorem hypothesis.

   Shown in the video
2. Expression

   $$Q \mapsto \lambda_1 y_1^2+\lambda_2 y_2^2+\cdots+\lambda_n y_n^2$$
   Explanation

   Make a change of variables that removes cross-product terms, leaving only squared terms weighted by eigenvalues.

   Justification

   Explicitly described by the lecturer as the essence of the proof.

   Shown in the video
3. Expression

   $$\lambda_i>0\ \forall i \Rightarrow Q>0 \text{ for } \vec{y}\neq \vec{0}$$
   Explanation

   If every coefficient is positive, each nonzero choice of the $y_i$ contributes positively, so the whole form is positive.

   Justification

   Spoken reasoning from the signs of the coefficients in the diagonal form.

   Shown in the video
4. Expression

   $$\lambda_i<0\ \forall i \Rightarrow Q<0 \text{ for } \vec{y}\neq \vec{0}$$
   Explanation

   If every coefficient is negative, every nonzero input gives a negative output.

   Justification

   Spoken reasoning from the signs of the coefficients in the diagonal form.

   Shown in the video
5. Expression

   $$\exists i,j:\lambda_i>0,\ \lambda_j<0 \Rightarrow Q \text{ takes both signs}$$
   Explanation

   With both positive and negative eigenvalues, one can obtain both positive and negative outputs, so Q is indefinite.

   Justification

   Spoken reasoning from the mixed-sign diagonal form.

   Shown in the video

The sign pattern of the eigenvalues of A determines whether Q is positive definite, negative definite, or indefinite.

1. Formula
   Observation

   The board writes $Q = x_1^2$.

2. Animation
   Observation

   A standard upward-opening parabola is drawn in an xy-plane.

3. Audio
   Observation

   The speaker says, "if I have the quadratic form Q equals x one squared... we can draw our xy plane and draw our graph as the parabola."

Graph the quadratic form $Q = x_1^2$ by treating it as a function.

1. $Q = x_1^2$

Identify the graph shape of the quadratic form.

1. Expression

   $$Q = x_1^2$$
   Explanation

   Start from the written one-variable quadratic form.

   Justification

   Directly shown on the board.

   Shown in the video
2. Expression

   $$f(x)=x^2$$
   Explanation

   Reinterpret the form as an ordinary single-variable function.

   Justification

   Spoken by the lecturer.

   Shown in the video
3. Expression

   $$\text{upward-opening parabola}$$
   Explanation

   Draw the familiar graph in the plane.

   Justification

   The speaker says this is how to graph it, and the animation shows a parabola.

   Shown in the video

The graph is an upward-opening parabola.

Later the lecturer classifies this example as positive definite because the parabola lies above the x-axis except at the origin.

1. Formula
   Observation

   The board writes $Q = -2x_1^2$.

2. Animation
   Observation

   A downward-opening parabola is drawn.

3. Audio
   Observation

   The speaker says, "If I have the quadratic form Q equals say negative two x one squared, then that's going to look like an inverted parabola."

Graph the quadratic form $Q = -2x_1^2$.

1. $Q = -2x_1^2$

Identify the graph shape after changing the coefficient sign.

1. Expression

   $$Q = -2x_1^2$$
   Explanation

   Use the written negative-coefficient quadratic form.

   Justification

   Directly shown on the board.

   Shown in the video
2. Expression

   $$\text{inverted parabola}$$
   Explanation

   The negative coefficient flips the parabola downward.

   Justification

   Stated verbally and illustrated by the drawing.

   Shown in the video

The graph is an inverted (downward-opening) parabola.

Later the lecturer classifies this example as negative definite because the parabola lies below the x-axis for nonzero inputs.

1. Formula
   Observation

   The board writes $Q = x_1^2 + x_2^2$.

2. Animation
   Observation

   A three-dimensional bowl-shaped surface is drawn.

3. Audio
   Observation

   The speaker says, "Now if I have a quadratic form with more variables, say Q equals x one squared plus x two squared, then I can draw this in three dimensions, and the shape that we have is a paraboloid."

Graph the two-variable quadratic form $Q = x_1^2 + x_2^2$.

1. $Q = x_1^2 + x_2^2$

Identify the three-dimensional surface shape.

1. Expression

   $$Q = x_1^2 + x_2^2$$
   Explanation

   Move from one variable to two variables.

   Justification

   The speaker explicitly introduces a form "with more variables."

   Shown in the video
2. Expression

   $$\text{paraboloid}$$
   Explanation

   Draw the surface in three dimensions.

   Justification

   The speaker names the shape as a paraboloid, and the animation shows a bowl-like surface.

   Shown in the video

The graph is a paraboloid.

Later the lecturer classifies this example as positive definite because the paraboloid lies above the xy-plane.

1. Formula
   Observation

   The board writes $Q = x_1^2 - x_2^2$.

2. Animation
   Observation

   A saddle-shaped surface is drawn.

3. Audio
   Observation

   The speaker says, "And if I have a quadratic form like Q equals x one squared minus x two squared, then we have a hyperbolic surface. What you may have learned in calculus is that these surfaces have saddle points in the middle."

Graph the two-variable quadratic form $Q = x_1^2 - x_2^2$.

1. $Q = x_1^2 - x_2^2$

Identify the surface type and its notable feature.

1. Expression

   $$Q = x_1^2 - x_2^2$$
   Explanation

   Change the plus sign in the previous two-variable example to a minus sign.

   Justification

   Directly shown on the board.

   Shown in the video
2. Expression

   $$\text{hyperbolic surface}$$
   Explanation

   The resulting graph is named a hyperbolic surface.

   Justification

   Stated verbally and illustrated by the drawing.

   Shown in the video
3. Expression

   $$\text{saddle point in the middle}$$
   Explanation

   The lecturer notes that such surfaces have saddle points at the center.

   Justification

   Explicitly said in the audio.

   Shown in the video

The graph is a hyperbolic surface with a saddle point in the middle.

Later the lecturer classifies this example as indefinite because some directions give positive outputs and others give negative outputs.

1. Audio
   Observation

   The speaker says, "So for example, Q equals x one squared here is positive definite because our parabola lies above the x axis. All of the outputs are positive whenever you input something that's not zero."

2. Audio
   Observation

   The speaker continues, "Our paraboloid here, Q equals x one squared plus x two squared, is also positive definite because our paraboloid is above the xy plane."

Determine which of the earlier graphed examples are positive definite.

1. $Q = x_1^2$

2. $Q = x_1^2 + x_2^2$

3. Definition: $Q(\vec{x}) > 0$ for all $\vec{x} \neq \vec{0}$

Apply the positive definite criterion to the examples.

1. Expression

   $$Q = x_1^2$$
   Explanation

   Check the one-variable example against the definition.

   Justification

   The example was already graphed earlier.

   Shown in the video
2. Expression

   $$\text{parabola lies above the x-axis}$$
   Explanation

   For nonzero inputs, the outputs are positive.

   Justification

   Stated verbally by the lecturer.

   Shown in the video
3. Expression

   $$Q = x_1^2 + x_2^2$$
   Explanation

   Check the two-variable example against the same definition.

   Justification

   The example was already graphed earlier.

   Shown in the video
4. Expression

   $$\text{paraboloid lies above the xy-plane}$$
   Explanation

   For nonzero inputs, the outputs are positive.

   Justification

   Stated verbally by the lecturer.

   Shown in the video

Both $Q = x_1^2$ and $Q = x_1^2 + x_2^2$ are positive definite.

Each example satisfies $Q(\vec{x}) > 0$ for all nonzero $\vec{x}$, as explained from the graphs.

1. Audio
   Observation

   The speaker says, "So for example, q equals negative two x one is negative definite because our parabola here lies below the x axis whenever your input is not zero."

1. The spoken phrase omits the square in "negative two x one," while the board consistently shows $Q = -2x_1^2$.

Determine whether $Q = -2x_1^2$ is negative definite.

1. $Q = -2x_1^2$

2. Definition: $Q(\vec{x}) < 0$ for all $\vec{x} \neq \vec{0}$

Apply the negative definite criterion to the example.

1. Expression

   $$Q = -2x_1^2$$
   Explanation

   Use the previously graphed negative-coefficient example.

   Justification

   The board shows this exact form.

   Shown in the video
2. Expression

   $$\text{parabola lies below the x-axis}$$
   Explanation

   For nonzero inputs, the outputs are negative.

   Justification

   Stated verbally by the lecturer.

   Shown in the video

$Q = -2x_1^2$ is negative definite.

It satisfies $Q(\vec{x}) < 0$ for all nonzero $\vec{x}$, as explained from the graph.

1. Audio
   Observation

   The speaker says, "So an example of that is our hyperbolic surface Q equals x one squared minus x two squared. On these tips, we have that the output is positive, while down here, the outputs are negative."

Determine whether $Q = x_1^2 - x_2^2$ is indefinite.

1. $Q = x_1^2 - x_2^2$

2. Definition: indefinite if $Q(\vec{x})$ has both positive and negative values

Apply the indefinite criterion to the saddle-surface example.

1. Expression

   $$Q = x_1^2 - x_2^2$$
   Explanation

   Use the previously graphed hyperbolic surface.

   Justification

   The board shows this exact form.

   Shown in the video
2. Expression

   $$\text{tips give positive output}$$
   Explanation

   Some directions on the surface produce positive values.

   Justification

   Stated verbally by the lecturer.

   Shown in the video
3. Expression

   $$\text{other region gives negative output}$$
   Explanation

   Other directions produce negative values.

   Justification

   Stated verbally by the lecturer.

   Shown in the video

$Q = x_1^2 - x_2^2$ is indefinite.

It has both positive and negative values, matching the definition given on the board.

1. Formula
   Observation

   The example box shows $Q = x_1^2 + 4x_1x_2 - 2x_2^2$.

2. Formula
   Observation

   The board writes A = [1 2; 2 -2] and $\lambda = -3$, 2.

3. Audio
   Observation

   The lecturer says this is the quadratic form from the last video, identifies the matrix and eigenvalues, and concludes that since A has both positive and negative eigenvalues, we can say ...

1. The final spoken sentence is cut off before the word 'indefinite' is heard.

2. The conclusion is strongly indicated by the theorem and the mixed eigenvalues, but the last word is not audible in this clip.

Revisit the quadratic form $Q=x_1^2+4x_1x_2-2x_2^2$ and classify it using its symmetric matrix and eigenvalues.

1. $Q=x_1^2+4x_1x_2-2x_2^2$.

2. $A=\begin{bmatrix}1&2\\2&-2\end{bmatrix}$.

3. $\lambda=-3$,\,2.

Determine the sign type of Q from the eigenvalues of A.

1. Expression

   $$Q=x_1^2+4x_1x_2-2x_2^2$$
   Explanation

   Write the quadratic form to be classified.

   Justification

   Shown in the example box.

   Shown in the video
2. Expression

   $$A=\begin{bmatrix}1&2\\2&-2\end{bmatrix}$$
   Explanation

   Use the symmetric matrix associated with Q.

   Justification

   Written on the board and stated in the audio.

   Shown in the video
3. Expression

   $$\lambda=-3,\,2$$
   Explanation

   Compute or recall the eigenvalues of A.

   Justification

   Written on the board and stated in the audio.

   Shown in the video
4. Expression

   $$\text{one eigenvalue}<0,\ \text{one eigenvalue}>0$$
   Explanation

   Observe that A has both a negative and a positive eigenvalue.

   Justification

   Direct comparison of the listed eigenvalues.

   Shown in the video
5. Expression

   $$Q \text{ is indefinite}$$
   Explanation

   Apply the theorem's indefinite criterion.

   Justification

   Derived from claim-eigenvalue-criterion; the final spoken word is truncated, so this conclusion is derived rather than fully heard.

   Derived from the video

Q is indefinite.

Check against the theorem: a symmetric matrix with both positive and negative eigenvalues yields an indefinite quadratic form. Here the eigenvalues are -3 and 2, so the criterion applies.

1. Formula
   Observation

   Example: $Q = x_1^2 + 4x_1x_2 - 2x_2^2$, $A = [[1, 2], [2, -2]]$, lambda = -5, 2, Q is indefinite!

Classify the quadratic form $Q = x_1^2 + 4x_1x_2 - 2x_2^2$.

1. $Q = x_1^2 + 4x_1x_2 - 2x_2^2$

Determine if Q is positive definite, negative definite, or indefinite.

1. Expression

   $$A = \begin{bmatrix} 1 & 2 \\ 2 & -2 \end{bmatrix}$$
   Explanation

   Identify the symmetric matrix A associated with the quadratic form.

   Justification

   Standard construction of the matrix from a quadratic form.

   Shown in the video
2. Expression

   $$\lambda = -5, 2$$
   Explanation

   Calculate the eigenvalues of matrix A.

   Justification

   Eigenvalue computation.

   Shown in the video
3. Expression

   $$Q \text{ is indefinite!}$$
   Explanation

   Conclude that Q is indefinite because the eigenvalues have mixed signs.

   Justification

   Applying the theorem relating eigenvalue signs to quadratic form classification.

   Shown in the video

Q is indefinite.

The eigenvalues are -5 and 2, which have opposite signs, confirming the indefinite classification according to the theorem.

1. Animation
   Observation

   Four handwritten examples are added left to right: $Q = x_1^2$ with an upward parabola, $Q = -2x_1^2$ with a downward parabola, $Q = x_1^2 + x_2^2$ with a bowl-shaped paraboloid, and $Q = x_1^2 - x_2^2$ with a saddle-shaped hyperbolic surface.

2. Audio
   Observation

   The narration names each shape as it is drawn: parabola, inverted parabola, paraboloid, and hyperbolic surface with saddle points.

1. $Q = x_1^2$

2. $Q = -2x_1^2$

3. $Q = x_1^2 + x_2^2$

4. $Q = x_1^2 - x_2^2$

5. xy-plane sketches

6. three-dimensional surface sketches

1. The first graph is an upward-opening parabola.

2. The second graph flips downward when the coefficient becomes negative.

3. The third graph moves from two dimensions to three dimensions and becomes a paraboloid.

4. The fourth graph changes the sign between squared terms and becomes a saddle-shaped hyperbolic surface.

1. All four objects are quadratic forms written as polynomials in $x_1$ and/or $x_2$.

2. The visual comparison keeps the algebraic formula above each corresponding sketch.

The animation links algebraic sign patterns to geometric shape types, preparing the later sign-based classification of quadratic forms.

1. Formula
   Observation

   A new block titled Definition appears with three bullets: positive definite, negative definite, and indefinite.

2. Animation
   Observation

   After the definitions are written, the lecturer visually returns attention to the earlier example graphs while explaining which ones satisfy each definition.

1. Definition heading

2. positive definite bullet

3. negative definite bullet

4. indefinite bullet

5. earlier four example graphs

1. The board shifts from example drawing to formal symbolic definitions.

2. The definitions are then applied back to the previously drawn graphs.

1. The same four example quadratic forms remain visible as reference cases.

2. The classification remains based on the sign of $Q(\vec{x})$.

The visual structure turns the earlier geometric intuition into explicit symbolic criteria for classifying quadratic forms.

1. Diagram
   Observation

   The whiteboard is organized into four zones: top examples with graphs, middle-left black definitions, middle-right blue semidefinite additions, and bottom theorem plus later example box.

1. Title 'Classifying Quadratic Forms:'

2. Four example formulas with sketches

3. Black definition block

4. Blue semidefinite block

5. Theorem block

6. Example box added at lower right

1. The board starts with examples and definitions already visible.

2. Blue semidefinite lines are added around 3s and 11s.

3. The theorem block appears around 30s.

4. The example box appears around 124s, followed by A and eigenvalues.

1. The top-row examples remain visible throughout.

2. The black definition block remains visible throughout once present.

The visual layout separates concrete examples, formal definitions, the main theorem, and a worked application.

1. Diagram
   Observation

   Top row shows $Q=x_1^2$ with upward parabola, $Q=-2x_1^2$ with downward parabola, $Q=x_1^2+x_2^2$ with upward paraboloid, and $Q=x_1^2-x_2^2$ with saddle surface.

1. $Q=x_1^2$ and upward-opening parabola

2. $Q=-2x_1^2$ and downward-opening parabola

3. $Q=x_1^2+x_2^2$ and bowl-shaped paraboloid

4. $Q=x_1^2-x_2^2$ and saddle surface

1. Each graph stays fixed once shown.

2. The purple curves/surfaces visually encode positivity, negativity, or mixed behavior.

The sketches provide geometric intuition: positive forms open upward, negative forms open downward, and indefinite forms have a saddle shape.

1. Animation
   Observation

   Blue handwritten text is added to the right of the black definition block.

2. Formula
   Observation

   First 'positive semidefinite if $Q(\vec{x}) \ge 0$.' then 'negative semidefinite if $Q(\vec{x}) \le 0$.'

1. Blue bullet for positive semidefinite

2. Blue bullet for negative semidefinite

1. At about 3s the positive semidefinite line appears.

2. At about 11s the negative semidefinite line appears below it.

1. The original black definitions remain unchanged while the blue extensions are added.

The blue text broadens the classification from strict definiteness to non-strict semidefiniteness.

1. Animation
   Observation

   A new theorem block appears beneath the definitions.

2. Formula
   Observation

   It states 'Theorem: If $Q = \vec{x}^T A \vec{x}$, then' followed by three iff bullets about eigenvalues of A.

1. Theorem heading

2. Condition $Q=\vec{x}^T A\vec{x}$

3. Three eigenvalue-based iff statements

1. The theorem block is absent before 30s and fully visible afterward.

1. Once written, the theorem remains on screen through the rest of the clip.

This is the central result of the segment: classification of Q by eigenvalue signs of A.

1. Animation
   Observation

   A boxed example appears at the lower right.

2. Formula
   Observation

   Inside the box: $Q = x_1^2 + 4x_1x_2 - 2x_2^2$, then A = [1 2; 2 -2], then $\lambda = -3$, 2.

1. Example label

2. Quadratic form $Q=x_1^2+4x_1x_2-2x_2^2$

3. Matrix A

4. Eigenvalue list

1. At about 124s the example box and Q appear.

2. At about 134s the matrix A is written.

3. At about 142s the eigenvalues are written.

1. The theorem and definitions stay visible while the example is built.

The example applies the theorem to a specific quadratic form by listing its matrix and eigenvalues.

1. Diagram
   Observation

   Four graphs are shown: $Q=x_1^2$ (parabola opening up), $Q=-2x_1^2$ (parabola opening down), $Q=x_1^2+x_2^2$ (bowl shape), $Q=x_1^2-x_2^2$ (saddle shape).

1. Graph of $Q=x_1^2$

2. Graph of $Q=-2x_1^2$

3. Graph of $Q=x_1^2+x_2^2$

4. Graph of $Q=x_1^2-x_2^2$

1. The shapes of the graphs correspond to the classifications of the quadratic forms.

The graphs visually demonstrate the concepts of positive definite (opening upwards/bowl), negative definite (opening downwards), and indefinite (saddle shape) quadratic forms.

1. Formula
   Observation

   The positive and negative definite bullets explicitly require the inequality only for $\vec{x} \neq \vec{0}$.

2. Audio
   Observation

   The speaker repeats "whenever x is not the zero vector" and "for all x's that's not the zero vector."

One might think the inequalities $Q(\vec{x})>0$ or $Q(\vec{x})<0$ must hold for every input including $\vec{x}=\vec{0}$.

The video states the conditions only for $\vec{x} \neq \vec{0}$; the zero vector is excluded from the definiteness test.

1. Formula
   Observation

   The board contrasts $Q = x_1^2 + x_2^2$ with $Q = x_1^2 - x_2^2$.

2. Audio
   Observation

   The speaker explains that the minus-sign version has positive outputs in some directions and negative outputs in others.