liveBook · Manning

Deceptive function

As explained in Optimization Algorithms: AI techniques for design, planning, and control problems

An objective-function landscape whose local structure can mislead search away from the global optimum.

A deceptive function has a narrow global minimum and a broad, strong local minimum with a similar value. The local basin can mislead search algorithms away from the best solution. section 2.1.2

Distinguish deceptive landscapes from related optimization cases:

  • Unimodal functions have one clear global optimum. section 2.1.2
  • Multimodal functions contain more than one optimum. section 2.1.2
  • Deceptive functions add a wide attractive basin around a near-global local minimum, increasing the risk of search stagnation. section 2.1.2

Visualizing a deceptive landscape

A Python objective_deceptive(x, y) function combines a narrow trigonometric structure with a broader quadratic term. Plotting its values as a 3D surface alongside unimodal and multimodal functions makes the misleading basin visible.

How Optimization Algorithms: AI techniques for design, planning, and control problems covers it

How deep it goes
Defines the landscape pattern and visualizes it computationally, but does not develop a specific remedy for deception.
Examples
Uses a Python three-dimensional surface plot comparing unimodal, multimodal, and deceptive objective functions.
What you can do after
Evaluate optimization landscapes by recognizing when a broad local basin can conceal a narrow global optimum.
How it connects
Connects deceptive functions with objective functions, global minima, local minima, unimodal functions, multimodal functions, and search algorithms.

This explainer was generated by Manning's knowledge graph from the text of Optimization Algorithms: AI techniques for design, planning, and control problems. The section links point to the parts of the book it draws on.