Heuristics Inspired by Nature – by Joe Wilke, Ph.D.
Business Analytics includes the use of algorithms and heuristics to solve complex problems. But have you ever stopped to wonder where those algorithms and heuristics come from? How do scientists and engineers contemplate such things? The simple answer … nature. The world around us is full of algorithmic inspiration. Here are some of my favorite examples:
- Steepest descent or steepest ascent is an algorithmic approach that finds the local minimum or maximum by looking for the steepest path. It is kind of like climbing a mountain or going down. It is the computational equivalent to being king of the hill.
- Simulated annealing is based on the annealing of metals. To make metals stronger they must be heated and cooled at precise intervals. This heuristic is used to avoid the local optimal solutions (e.g., a false peak) and take us to the true maximum (or minimum) of the solution space.
- Genetic algorithms are based on evolution and the survival of the fittest. Imagine being able to create a large population of possible solutions and then have them mix-and-mingle-and-mutate to achieve stronger future generations of solutions. After the creation of many generations and the survival of the fittest mentality, then you are left with an excellent solution to your overall problem. These algorithms are used in a variety of industries, recently they were used to optimize the shape of a special antenna for NASA ST5 spacecraft.
- Particle swarm and ant-colony optimization mimic animal behavior. Particle swarm mimics the way birds fly together or a school of fish move throughout the sea. Whereas, ant-colony optimization mimics how ants use pheromones to optimize their path from a food source to their colony. These types of algorithms are especially powerful for vehicle routing and social network modeling.
So, be on the lookout, using a heuristic or algorithm inspired by nature may be your best option when solving your next complex business problem.