carrot carrot carrot Change Centers x cognizanti collaborators create-folder Data Science Decisive Infrastructure download download edit Email exit Facebook files folders future-of-work global sourcing industry info infographic linkedin location Mass Empowerment Mobile First our-latest-thinking pdf question-mark icon_rss save-article search-article search-folders settings icon_share smart-search Smart Sourcing icon_star Twitter Value Webs Virtual Capital workplace Artboard 1

No Results.

Did you mean...

Or try searching another term.

Evolutionary Computation/Evolutionary AI

What is evolutionary computation/evolutionary AI?

Evolutionary computation (EC) is inspired by natural evolution. In contrast to most techniques in engineering and design, where humans come up with the best solution possible, debug it, and deploy it, evolutionary computation provides a way of coming up with new, creative solutions automatically—often solutions that are too complex or unusual for humans to discover.

With evolutionary computation, human engineers define how the quality of potential solutions is measured (i.e. the fitness) and what kind of solutions are possible (i.e. the search space). Given such definitions, any engineering, healthcare, life science, insurance or business problem can be solved using evolutionary computation. Examples include neural network architectures, growth recipes for plants, electric circuits, or even football schedules. Among the possible solutions, evolutionary computation will home in on those with the best quality. Evolutionary computation is ideal for any complex analysis where it is not possible for people to evaluate all the variable interactions in a timely manner.

The process begins with an initial population of possible solutions, as widely different as possible. Each solution is evaluated, and the population rank ordered. The least useful solutions are removed and replaced with offspring of the most useful solutions, formed by mutating and recombining them. The evaluation-selection-recombination loop repeats, gradually creating better candidates, until satisfactory solutions are found.

What are the business benefits of evolutionary computation/evolutionary AI?

Businesses aiming to improve designs or processes are understandably overwhelmed when the number of variables grows beyond about 15. These variables often interact and create a search space of millions of possible solutions, or more. In these cases, it is no longer possible for humans to handle this many variables but evolutionary computation is a tireless solution-generating engine that can.

Oftentimes the results found by evolutionary computation are not only viable but also surprising. Note that these are not engineered or designed by humans. Yet, they provide unique solutions to problems that are defined by humans, and therefore extend their problem-solving ability. Cognizant’s patented evolutionary computation solution is powerful enough to solve some of the world's most complex business problems. It is also inherently parallelizable, making it possible to evaluate millions of potential solutions over thousands of CPUs, which means it can be scaled to very large problems.