Artificial Neural Network

Artificial Neural Network

Introduction: 

“A computer system modeled on human brain or nervous system”- Wikipedia

The inventor of first neuro-computer Dr. Robert Hecht- Neilson defines a neural network as – 

“… a computing system made up of a number of simple highly interconnected processing elements which process information by their dynamic state response to external inputs”

An artificial neural network usually involves a large number of processors operating in parallel and arranged in tiers. The first tier receives the raw input information analogous to optic nerves in human vision processing. Each successive tier receives the output of preceding tier, rather than the raw input. The last tier produces output of the system.

Each processing node has its own small sphere of knowledge, including what it has seen and any rules it was originally programmed with, or developed for itself.

Definition: 

What is neural network? 

A neural network is an artificial representation of the human brain that tries to simulate its learning process. An artificial neural network is often called as Neural Network or Neural Net

Traditionally, the word neural network is referred to a network of biological neurons in nervous system that process and transmit the information. 

ANN is an interconnected group of artificial neurons that uses a mathematical model or computational model for information processing based on a connectionist approach to computation.

The artificial neurons may share some properties of biological neurons

ANN is an adaptive system that changes its structure based on external or internal information that flows through the network.

Neural computers mimic certain processing capabilities of human brain 

Neural computing is an information processing paradigm inspired by biological system , composed of large number if highly interconnected processing elements (neurons) working in union to solve specific problem.

An ANN is configured for a specific application such as pattern recognition, or data classification through a learning process.

A simple artificial neuron

• Takes input
• Calculate summation of inputs
• Compare it with threshold being set during learning stage
• Outputs the result 

 ANNs are non linear data driven self adaptive approach as opposed to traditional model based methods. They are powerful tools for modelling especially when relation between the underlying data is unknown.

ANNs can identify and learn correlated patterns between input data sets and corresponding target values. 

After training ANNs can be used to predict the outcome of new independent input data.

ANNs can imitate the learning process of human brain and can process problems involving non linear and complex data even if data is imprecise and noisy

A very important feature of ANN is their adaptive nature where learning by example replaces programming in solving problems

This feature makes computational models very appealing in application domains where one has little or incomplete understanding problem to be solved but where training data is readily available. These networks are “neural” in the sense that they are inspired by the neuroscience but not necessarily faithful models of biological neural phenomena. 

In fact majority of network are more closely related to traditional mathematical and/ or statistical models such as Non- Parametric Pattern classifiers, clustering algorithms, non linear filters, statistical regression models than they are to neurobiology models. 

NNs have been used in wide range of applications. They have been used for classification problems such as identifying underwater current sonar, recognizing speech and predicting the secondary structure of globular proteins. NNs have also been used in predicting stock market performance.

Characteristics of Neural Network:

1.                   The NNs exhibit mapping capabilities, that is, they can map input patterns to their associated output patterns.

2.                   NNs learn by example. Thus NN architectures can be trained with known examples of a problem before they are tested for their inference capabilities on unknown instances of problem. They can, therefore identify new objects previously untrained. 

3.                    The NNs posses the capability to generalize. Thus hey can predict new outcomes from past trends. 

4.                   The NNs are robust system and are fault tolerant. They

 can therefore recall full patterns from incomplete , partial or noisy patterns 

5.                   NNs can process information in parallel, at high speed, in a distributed manner.