Study of Artificial Neural Network Sunakshi Jain#1, Yatika Goel#2 , Deepali Singhal#3 #1,2 Computer Science & Engineering Department, R.K.G.I.T.W. U.P. Technical University, U.P. 1swtsunakshi19@gmail.com 2yatikag92@gmail.com 3deepalisinghal@rkgitw.edu.in Abstract— In this paper, we are elaborating Artificial Neural Network or ANN, its various characteristics and business applications. In this paper we also show that “what are neural networks” and “Why they are so important in today’s Artificial intelligence?” ANN provides a very exciting alternatives and other application which can play important role in today’s computer science field. The concept of ANN is basically introduced from the subject of biology where neural network plays a important and key role in human body. In human body work is done with the help of neural network. There are some Limitations also which are mentioned. Keywords— Artificial Neural Network, Artificial Neuron, Characteristics, Applications, Limitations. 3. channels, all inputs from other neurons arrive through the dendrites. Axon- It also attached to the soma and is electrically active and serves as an output channel. The structural constituents of a human brain termed Neurons are the entities, which perform computations such as cognition, logical inference, and pattern recognition and so on. Hence the technology, which has been built on a simplified imitation of computing by neurons of a brain, has been termed Artificial Neural Systems (ANS) technology or Artificial Neural Network (ANN). I. Neural network architecture I. Introduction Work on artificial neural networks commonly referred to as “neural networks”, has been motivated right from its inception by the recognition that the human brain computers in an entirely different way from the conventional digital computers. The human brain is one of the most complicated things .Brain contains about 1010 basic units called Neurons. A neuron is a small cell that receives electro-chemical signals from its various sources and in turn responds by transmitting electrical impulses to other neurons. Despite their different activities, all neurons share common characteristics:1. Soma – A neuron is composed of a nucleus-a cell body known as soma. 2. Dendrite – Attached to the soma are long irregularly shaped filaments called dendrites. It behaves as input An Artificial Neural Network is defined as a data processing system consisting of a large number of simply highly interconnected processing elements in an architecture inspired by the structure of the cerebral cortex of the brain. There are several classes of NN, classified according to their learning mechanisms. However, we identify three fundamentally different classes of Networks. Similar to biological Neuron Artificial Neural Network also have neurons which are artificial and they also receive inputs from the other elements or other artificial neurons and then after the inputs are weighted and added, the result is then transformed by a transfer function into the output. The transfer function may be anything like Sigmoid, hyperbolic tangent functions or a step. Even though it has its own limitations, it is applied to a wide a range of practical problems and has successfully demonstrated its power. III. Characteristics of artificial neural network Conventionally, a computer operates through sequential linear processing technologies. They apply formulas, decision rules, and algorithms instructed by users to produce outputs from the inputs. Conventional computers are good at numerical computation. But ANNs improve their own rules; the more decisions they make, the better the decisions may become. There are six main characteristics of ANN technology Network Structure-The Network Structure of ANN should be simple and easy. There are basically two types of structures recurrent and non recurrent structure. The Recurrent Structure is also known as Auto associative or Feedback Network and the Non Recurrent Structure is also known as Associative or feed forward Network. Parallel Processing Ability- Parallel Processing is done by the human body in human neurons are very complex but by applying basic and simple parallel processing techniques we implement it in ANN like Matrix and some matrix calculations. Distributed Memory-ANN is very huge system so single place memory or centralized memory cannot fulfil the need of ANN system so in this condition we need to store information in weight matrix which is form of long term memory because information is stored as patterns throughout the network structure. Fault Tolerance Ability-ANN is a very complex system so it is necessary that it should be a fault tolerant. Because if any part becomes fail it will not affect the system . Learning Ability-In ANN most of the learning rules are used to develop models of processes, while adopting the network to the changing environment and discovering useful knowledge. IV. BACKPROPAGATION NETWORKS Back propagation is a systematic method of training multilayer artificial neural networks. It is built on high mathematical foundation and has very good application potential. A. Single layer artificial neural network Consider a single layer feed forward neural network consisting of an input layer to receive the inputs and an output layer to output the vectors respectively. The input layer consists of ‘n’ neurons and the output layer consists of ’m’ neurons. Indicate the weight of the synapse connecting ith input neuron to the jth output neuron as wij . The inputs of the input layer and the corresponding outputs of the output layer are given as Assume, we use linear transfer function for the neurons in the input layer and the unipolar sigmoidal function for the neurons in the output layer OI = I I (linear transfer function) nx1 mx1 IOj =W1jII1 + W2jII2 +…………….+WnjIIN Hence , the input to the output layer can be given as O IO = [W]T OI = [W]T I I mx1 mxn nx1 Using the unipolar sigmoidal or squashed-S function and the slope of this function for neurons in the output layer, the output is given by OOk = 1/(1 + e-λIOk) OO = f[WI] B. Model for multilayer perceptron The adapted perceptrons are arranged in a layers and so the model is termed as multilayer perceptron. This model has three layers: an input layer, and output layer, and a layer in between not connected directly to the input or the output and hence, called the hidden layer. For the perceptrons in the input layer, we use linear transfer function, and for the perceptrons in the hidden layer and the output layer, we use sigmoidal or squashed-S functions. The input-output mapping of multilayer perceptron is represented by O=N3[N2[N1[I]]] N1, N2, and N3 represent nonlinear mapping provided by input, hidden and output layers respectively. Multilayer perceptron provides no increase in computational power over a single layer neural network unless there is a nonlinear activation function between layers. V. Applications of artificial neural network There are many applications of ANNs in today's business. Financial institutions are improving their decision making by enhancing the interpretation of behavioral scoring systems and developing superior ANN models of credit card risk and bankruptcy. Two actual ANN applications are now described. A. Airline Security Control With the increasing threat of terrorism, airline passengers' bags in international airports such as New York and London go through an unusually rigorous inspection before being loaded into the cargo bay. In addition to using metal detector and x-ray station to detect metal weapons, these airports use ANNs to screen for plastic explosives. They use a detection system which bombards the luggage with neutrons and monitors the gamma rays that are emitted in response. The network then analyzes the signal to decide whether the response predicts an explosive. Explosive materials are rich in nitrogen, but so are some benign substances, including protein-rich materials, such as wool and leather. To minimize the classification error, supervised training was conducted. The ANN was fed with a batch of instrument reading as well as the information on whether explosives were indeed present. The trained network was able to achieve its intended purpose. The reduction in false alarms translates into many less bags that must be opened and examined each day. In turn, it reduces the cost of airport operations, increases the efficiency of the check-in process, and improves the satisfaction of Customers. B. Investment Management and Risk Control Neural Systems makes use of a supervised network to mimic the recommendations of money managers on the optimal allocation of assets among Treasury instruments. The application demonstrated how well an ANN can be trained to recognize the shape and evolution of the interest yield curve. The network was trained on measured and calculated economic indicators, such as the evolution of interest rates, price changes, and the shape and speed of the change of the yields curves. The network could then determine the optimal allocation among segments in various Treasury instruments being measured against a benchmark or comparator performance index. It could determine also the dynamic relationship between different variables in portfolio management and risk control. each layer, the transfer function of each neuron, and the size of the training set. While too small a training set will prohibit the network from developing generalized patterns of the inputs, too large a one will break down the generalized patterns and make the network sensitive to input noise. This calls for easyto use and easy-to-modify ANN development tools that are gradually appearing on the market. C. The output quality of an ANN may be unpredictableThis may not be the case for finding the solution to a problem with linear constraints in which the solution, if found, is guaranteed to be the global optimum. A solution to a non-linear problem reached by the ANN may not be the global optimum. In a mission-critical application, one should develop ANN solutions in parallel with the conventional ones for direct comparison. Both types of systems should be run for a period of time, long enough to make sure that the ANN systems are error-free before they are used in real situations. VII. CONCLUSIONS The field of ANN went through a dormant period, because the early single layer models were fundamentally flawed. Soon after, some multi-layer and trainable ANN models emerged. Despite having some inherent limitations, ANNs have been increasingly popular since then. They are feasible for those business applications which require the solution of very complex system of equations, recognizing patterns from imperfect inputs, and adapting decisions to changing environment. Parallel Processing is more needed in this present time because with the help of parallel processing only we can save more and more time and money in any work related to computers and robots. VI. Limitations of Artificial Neural Networks REFERENCES Artificial neural network is undoubtedly a powerful tool for decision making. But there are several weaknesses in its use. A. ANN is not a general-purpose problem solver- It is good at complex numerical computation for the purposes of solving system of linear or non-linear equations, organizing data into equivalent classes, and adapting the solution model to environmental changes. However, it is not good at such tasks as calculating payroll, balancing checks, and generating invoices. B. There is no structured methodology available-No structured methodology available for choosing, developing, training, and verifying an ANN. The solution quality of an ANN is known to be affected by the number of layers, the number of neurons at [1] Aeh, R.K. "Connectionism," Journal of System Management, April 1989, p. 23. [2] Babloyantz, A. and Destexhe, A. "Chaos in Neural Networks," In M. Caudill and C. Butler (Eds.) IEEE First International Conference on Neural Networks (Vol. 4). Piscataway, NJ: IEEE Service Center, pp. 31-39. [3] Bernardino, L.A. "Untangling the Mystery of Neural Networks," Machine Design, Vol. 65, No. 13, June 25, 1992, pp. 55-59. [4]Ajith Abraham, “Artificial Neural Networks”, Stillwater,OK, USA, 2005. [5] Carlos Gershenson, “Artificial Neural Networks for Beginners”, United kingdom. [6] Anil K Jain, Jianchang Mao and K.M Mohiuddin, “Artificial Neural Networks: A Tutorial”, Michigan State university,1996. [7]Vincent Cheung and Kevin Cannons, “ An Introduction of Neural Networks”, Manitoba, Canada, May 27, 2002. [8]Howard Demuth and Mark Beale, “ Neural Network Toolbox”, With the help of metlab, user guide version 4. [9]Girish Kumar Jha, “Artificial Neural Network and its Applications”,IARI New delhi. [10] About Neural Network from website http://en.wikipedia.org / wiki/Neural_network .