Random search

In mathematical and numerical optimization, random search (RS) refers to a family of methods that do not require the gradient of the problem to be optimized and RS can hence be used on functions that are not continuous or differentiable. Such optimization methods are also known as direct-search, derivative-free, or black-box methods.

The name, random search, is attributed to Rastrigin who made an early presentation of RS along with basic mathematical analysis.

Algorithm

Let $f:\mathbb {R} ^{n}\rightarrow \mathbb {R}$ be the fitness or cost function which must be minimized. Let $\mathbf {x} \in \mathbb {R} ^{n}$ designate a position or candidate solution in the search-space. The basic RS algorithm can then be described as:

• Initialize $\mathbf {x}$ with a random position in the search-space.
• Until a termination criterion is met (e.g. number of iterations performed, or adequate fitness reached), repeat the following:
  • Sample a new position $\mathbf {y}$ from the hypersphere of a given radius surrounding the current position $\mathbf {x}$ (see e.g. Marsaglia's technique for sampling a hypersphere.)
  • If $(f(\mathbf {y} )<f(\mathbf {x} ))$ then move to the new position by setting $\mathbf {x} =\mathbf {y}$
• Now $\mathbf {x}$ holds the best-found position.

Variants

A number of RS variants have been introduced in the literature:

• Fixed Step Size Random Search (FSSRS) is Rastrigin's basic algorithm which samples from a hypersphere of fixed radius.
• Optimum Step Size Random Search (OSSRS) by Schumer and Steiglitz is primarily a theoretical study on how to optimally adjust the radius of the hypersphere so as to allow for speedy convergence to the optimum. An actual implementation of the OSSRS needs to approximate this optimal radius by repeated sampling and is therefore expensive to execute.
• Adaptive Step Size Random Search (ASSRS) by Schumer and Steiglitz attempts to heuristically adapt the hypersphere's radius. The algorithm, however, is somewhat complicated.
• Optimized Relative Step Size Random Search (ORSSRS) by Schrack and Choit approximate the optimal step size by a simple exponential decrease. Again, however, the formula for computing the decrease-factor is somewhat complicated.

See also

• Random optimization is a closely related family of optimization methods which sample from a normal distribution instead of a hypersphere.
• Pattern search takes steps along the axes of the search-space using exponentially decreasing step sizes.