Study Guide for Quiz 8 1. Eigenvalues Given a 2 × 2 matrix A, its eigenvalues λ and µ satisfy the equations: λ + µ = trace(A) and λµ = det(A). You can use this to find the eigenvalues of any 2 × 2 matrix. 2. The Matrix for a Quadratic Form Any quadratic form f (x, y) = ax2 + bxy + cy 2 on R2 can be written using a matrix: #" # " a b/2 x . f (x, y) = x y y b/2 c Similarly, a quadratic form f (x, y, z) = ax2 + by 2 + cz 2 + pxy + qxz + ryz can be written using a matrix: a p/2 q/2 x f (x, y) = x y z p/2 b r/2 y . q/2 r/2 c z 3. Classifying Quadratic Forms You can use the eigenvalues of the matrix to determine the shape of the graph: • If both eigenvalues are positive, the graph is an elliptic paraboloid that opens up. • If both eigenvalues are negative, the graph is an elliptic paraboloid that opens down. • If one eigenvalue is positive and the other is negative, the graph is a saddle surface. In the first two cases, the level curves for f are ellipses, while in the third case they are hyperbolas. 4. Definite Forms A quadratic form is positive definite if all of its eigenvalues are positive, and negative definite if all of its eigenvalues are negative. According to Sylvester’s criterion, a quadratic form a r s x y f (x, y, z) = x y z r b t s t c z is positive definite if and only if a>0 and a r r b > 0 and a r s r b t > 0. s t c You can also use this to check whether a quadratic form f : R3 → R is negative definite: just check whether the negation −f is positive definite. 5. Critical Points A critical point for a function f : R2 → R is a point for a ∈ R2 for which ∂f (a) = 0 ∂x ∂f (a) = 0. ∂y and More generally, a critical point for a function f : Rn → R is a point for a ∈ Rn for which ∂f (a) = 0, ∂x1 ∂f (a) = 0, ∂x2 ..., and ∂f (a) = 0. ∂xn To find the critical points for a function f , set all of the partial derivatives equal to zero and then solve the resulting system of equations. 6. The Hessian Matrix The Hessian of a function f : R2 → R is the following 2 × 2 matrix: 2 ∂ f ∂ 2f ∂x2 ∂x∂y Hf = ∂ 2f ∂ 2f ∂x∂y ∂y 2 Similarly, the Hessian of a function f : R3 → R is the following 3 × 3 matrix: 2 ∂ f ∂ 2f ∂ 2f ∂x2 ∂x∂y ∂x∂z 2 2 2 ∂ f ∂ f ∂ f Hf = ∂x∂y ∂y 2 ∂y∂z 2 2 ∂ 2f ∂ f ∂ f ∂x∂z ∂y∂z ∂z 2 The Hessian is a multivariable analogue of the second derivative of a function. We use the notation Hf (a) to denote the value of the Hessian matrix at a point a. 7. The Second Derivative Test Let f : Rn → R be a differentiable function, and let a be a critical point of f . Then: (a) If the Hessian Hf (a) is positive definite, then f has a local minimum at a. (b) If the Hessian Hf (a) is negative definite, then f has a local maximum at a. (c) If the Hessian Hf (a) has both negative and positive eigenvalues, then f has neither a local minimum nor a local maximum at a. 8. Multivariable Taylor Series The Taylor series for a function f : R2 → R has the form ∞ X ∞ X m=0 n=0 cm,n xm y n where cm,n = 1 ∂ m+n f (0). m! n! ∂xm ∂y n The second-order Taylor polynomial p2 (x, y) is the sum of the constant, linear, and quadratic terms of this Taylor series: p2 (x, y) = f (0) + ∂f ∂f 1 ∂ 2f 1 ∂ 2f ∂ 2f 2 (0) x + (0) y + (0) xy + (0) x + (0) y 2 ∂x ∂y 2 ∂x2 ∂x∂y 2 ∂y 2 This can also be written p2 (x) = f (0) + Df (0) x + 1 T x Hf (0) x. 2