The Lagrange Multiplier λ
You'll learn: How the Lagrange multiplier λ quantifies the relationship between the gradients of the objective function and the constraint at optimal points.
What Does λ Tell Us?
In the previous lesson, you learned that at an optimal point of a constrained problem, the gradients ∇f and ∇g must be parallel. The Lagrange multiplier λ (lambda) is the scalar that precisely captures this parallel relationship.
At an optimal point, we have:
∇f = λ∇g
This equation means the gradient of your objective function f (what you're optimizing) is a scalar multiple of the gradient of your constraint g (your boundary condition). The multiplier λ is that scaling factor.
Understanding λ
Think of λ as a "conversion rate" between the two gradients. If λ = 2, it means at the optimal point, the rate of change of f is exactly twice the rate of change of the constraint g in corresponding directions.
- Positive λ: Both gradients point in the same general direction (local maximum along the constraint)
- Negative λ: Gradients point in opposite directions (often a local minimum)
- Magnitude of λ: Tells you how sensitive your optimal value is to small changes in the constraint
For example, if you're maximizing profit f subject to a budget constraint g, then λ represents approximately how much additional profit you'd gain per unit increase in budget.
The Method's Core
Instead of solving the constrained problem directly, we introduce λ as an unknown and solve the system: ∇f = λ∇g along with the constraint equation g = c simultaneously.
Key Takeaway: The Lagrange multiplier λ is the scalar that makes ∇f = λ∇g true at optimal points, quantifying how the gradients relate and revealing the sensitivity of the optimum to constraint changes.