Boundary Considerations for Optimization
What You'll Learn
You'll learn why finding critical points inside a region isn't enough—you must also check the boundary to guarantee you've found the true maximum or minimum.
The Interior vs. The Edge
When optimizing a multivariable function over a closed and bounded region, the Second Derivative Test helps you classify critical points inside the region (the interior). These are points where both partial derivatives equal zero.
But here's the catch: The absolute maximum or minimum might not occur at an interior critical point at all—it could be sitting right on the boundary!
Think of it like searching for the highest point in a fenced park. You might find a hill inside the park (an interior critical point), but the actual highest elevation could be along the fence line itself, where the terrain slopes upward as it meets the boundary.
Why Boundaries Need Separate Analysis
Critical point analysis assumes you can move freely in all directions—that's how we set ∇f = 0. But on a boundary, you're constrained. You can't test whether moving "outside" the region would increase or decrease the function value, because you're not allowed to leave!
Example: Optimizing f(x, y) over the disk x² + y² ≤ 1
- Find interior critical points where fₓ = 0 and f_y = 0
- Then separately examine the boundary circle x² + y² = 1 (often using substitution or Lagrange multipliers, which you'll learn later)
- Compare all candidate values to determine the absolute extrema
The Complete Optimization Procedure
- Find all interior critical points
- Classify them using the Second Derivative Test
- Examine the boundary separately (parametrize it, substitute constraints, etc.)
- Compare function values at all candidates
Key Takeaway: Critical point analysis only finds interior extrema; to find absolute extrema on a closed region, you must examine boundary points separately and compare all candidates.