A concept exploding in the field of computation and artificial intelligence today is the Artificial Neural Network (ANN). The discovery is changing the way scientists think of computers and the way industrialists and medical professionals, among others, are using them.

Ranking: Top 30 Best Online Schools for Computer Science Degrees

Defining ANN

One of the pioneers in this concept is Dr. Robert Hecht-Nielson. He defines ANN as “a computing system made up of simple, highly interconnected processing elements, which process information by their dynamic state response to external inputs.” Simplified, ANN is a computer system made up of small processors that work together somewhat like the neurons in the human brain to understand the information presented to it. While the ANN system contains hundreds of thousands of these processors, or artificial neurons, the brain has billions. Each neuron receives information and interprets it according to its own function, then sends the information to other neurons that add their own interpretations until the system recognizes a pattern. For instance, a human child might see a poodle and be told it is a dog. When she sees a husky, she is told that it is a dog too. The child’s brain makes note of similarities. After receiving several of these messages, the child can differentiate a dog from other animals.

Differences from Other Computer Systems

The computing system most people are familiar with consists of a very complex central processor. It stores information in a huge “library” and retrieves it to solve problems. Algorithms are fed to the machine that gives it a pattern of work to perform like steps in a recipe. In the ANN system, each artificial neuron is actually a small processor that concentrates on the relevance of one aspect of the input information and then communicates its results to other neurons. The neuron may be an algorithm or even actual hardware. Because of the volume of “processors” present, the input is large, and the machine can make inferences from it. In a sense, it may sometimes call a whale a dog because it is a mammal. The differences between the dog and the whale become apparent as the other neurons share their information, and the system learns what is, and what is not, a dog. According to an article in the New York Times, conventional computational systems learn from the “top down” by applying rules to data. The vision for artificial intelligence is that it would learn from the “ground up” by interpreting data. The ANN concept is foundational to that vision.

How it is Used

Google is a front-runner in both the exploration and the application of artificial intelligence. The company has made the terms “enhanced reality” and “virtual reality” part of the common vernacular. The Google Translator has used the artificial neurons to move from merely translating a set of words to imparting the meaning of a phrase. Google Glasses allow people to look at historical buildings and be given the history and the context of the edifice. The New York Times likens the idea to giving a computer eyes, “building around capabilities that now exist to understand photos. Robots will be drastically transformed.” Medicine will see advances in diagnostics. Shipping companies will use the technology to find the shortest routes through traffic and construction. Car manufacturers will refine the capabilities of autonomous-driving cars.

Artificial intelligence is a controversial field. People speculate about machines that eventually make humans irrelevant. In reality, they are “art imitating reality.” The truth is that the Artificial Neural Network is a simple replication of a complex system.