At WiseGEEK, we're committed to delivering accurate, trustworthy information. Our expert-authored content is rigorously fact-checked and sourced from credible authorities. Discover how we uphold the highest standards in providing you with reliable knowledge.

Learn more...

What is a Neural Algorithm?

A. Leverkuhn
A. Leverkuhn

A neural algorithm commonly refers to a piece of code used in neural programming. This is where a neural network simulates specific behaviors and attributes of the human brain. Programmers talk about neural programming as a process evolved from older systems, where today's neural programming community builds on principles of artificial intelligence presented decades ago.

The neural algorithm is a specific part of neural systems that helps facilitate one of the greater roles of neural software. It often provides for combining different data for a specialized result, where the neural algorithm fills in the gaps much like a human cerebral process would do, for example, in a limited range of vision. In artificial neural programming, this is done by projecting from known data to present a likely result.

Neural algorithms simulates specific behaviors and attributes of the human brain.
Neural algorithms simulates specific behaviors and attributes of the human brain.

Many neural algorithm setups involve taking a known input, and adding another kind of "training data" to get a final result that combines both. Developers look closely at machine learning to define how well their neural algorithms are producing the capability of a computer program to learn. Beyond this, there are a wide range of types of neural algorithms intended for different goals and implemented in different ways.

Programmers often include detailed diagrams to show how each component of a neural algorithm blends into the mix. These may be published in print or on the web to help a public developer community interpret what a single programmer or team has done with a neural algorithm to enhance a piece of software. Like all programming, neural algorithm development relies heavily on conventional language and coding, standard documentation practices, and clarity from the original team to make the result accessible to a wider audience. Without this, it becomes difficult to translate the original intent and functionality of an algorithm or program.

Along with fundamental roles in fields like logistics and observational sciences, neural applications have now become popular in unlikely places. One of these is in horse racing, where developers of computer programs now claim that neural algorithms can be used to effectively predict outcomes. Although these kinds of uses are similar to other common practices for neural software design, it's debatable how well neural applications can predict a particular event. The interest in using a neural algorithm design to track data-rich events like stock market changes is big enough to ensure that neural programming will be a big part of future efforts to develop computer programs that help human operators in specific predictive ways.

Discussion Comments


@hamje32 - I find the use of the human brain as a model in computer programming to be fascinating in itself.

On one level it proves just how advanced the human brain is; I think the analogy between the brain and a CPU is a fair one. On another level, it proves how primitive even the most advanced computer technology is when put beside it.

Biological neural networks are lightning fast, and the ability to “learn” is not one that comes easily even to the most advanced computer programs, from what I understand.

Sure, they’ve built computers that can beat world champion chess players, but I think it will be years before computers can learn on a scale that the brain can, when compared with even the smallest brain in a human baby.


@everetra - I agree, there’s no magic bullet in stock market investing. You just have to use your best judgment and diversify as always.

But you can use the software to point out trends across a multitude of industries and stocks; you may not have the time to do all that kind of research on your own.

I would think it would use something like neural networks pattern recognition to recognize these trends and provide you with at least a set of options. You can make the final decision on your own.


@nony - I have yet to use a single investment software package that was as accurate as it claimed in predicting market outcomes.

Yes, I’ve tried those that used neural network learning – or at least claimed to, which I guess is a fancy way of saying that the software "learns" on its own or something like that.

The problem with some of these investment software packages is that they can’t account for the “X” factor – an unknown variable, like actions by the Federal Reserve, crises around the world, or sudden currency devaluations.

They are good at interpreting trend lines and looking at PE ratios and valuations; but you don’t need software to do that. You can study the charts and read the numbers for yourself.


I had a professor in college who wrote extensively on neural networks and genetic algorithms. I have to admit that his classes were quite challenging, but he was interesting to listen to.

He didn’t just write; he coded as well. He worked on one project where he was trying to use a genetic algorithm to solve this really advanced problem in computer science.

I don’t remember if this was for personal research or something he did for the military. However, whatever it was it required a lot of computer power. He had a whole bunch of Sun workstations connected together in a distributed computing format, working on the same problem for twenty four hours a day for several weeks. I

t was an ambitious endeavor but he eventually solved it.

Post your comments
Forgot password?
    • Neural algorithms simulates specific behaviors and attributes of the human brain.
      By: Alexandr Mitiuc
      Neural algorithms simulates specific behaviors and attributes of the human brain.