Britannica
Sources used in this answer
How does a local optimum differ from a global optimum in nonlinear optimization?
Created from Britannica articles
using AI.
AI can make mistakes, so verify using Britannica articles.
optimization
Constraint set bounded by the five lines x1 = 0, x2 = 0, x1 = 8, x2 = 5, and x1 + x2 = 10. These enclose an infinite number of points that represent feasible solutions.
In nonlinear optimization, the distinction between a local optimum and a global optimum is fundamental to how algorithms navigate the search space.
- Global Optimum: This represents the absolute best value of the objective function across the entire set of feasible solutions. Identifying this point is the ultimate goal of optimization, though it is often computationally impractical to guarantee in complex nonlinear problems.1
- Local Optimum: This is defined as the best value within a specific, restricted region of the feasible solutions. Because nonlinear functions can have multiple peaks and valleys, an algorithm may become trapped in a local optimum, appearing to have reached the best solution when a superior one exists elsewhere in the broader search space.1
Most nonlinear programming algorithms function by generating a sequence of guesses, or iterates, to improve the objective function value. Because these methods rely on local information gathered at each step, they are frequently targeted toward finding a local optimum. In many real-world applications, settling for a local optimum is a necessary compromise when the global optimum cannot be practically identified.1