Artificial Neural Network Artificial Neura ra

CHAPTER 2 LITERATURE REVIEW

2.1 Artificial Neural Network

Artificial neural network ANN, usually called Neural Network NN, is an algorithm that was originally motivated by the goal of having machines that can mimic the brain. A neural network consists of an interconnected group of artificial neurons. They are physical cellular systems capable of obtaining, storing information, and using experiential knowledge. Like human brain, the ANN’s knowledge comes from examples that they encounter. In human neural system, learning process includes the modifications to the synaptic connections between the neurons. In a similar way, ANNs adjust their structure based on output and input information that flows through the network during the learning phase. Data processing procedure in any typical neural network has two major steps: the learning and application step. At the first step, a training database or historical price data is needed to train the networks. This dataset includes an input vector and a known output vector. Each one of the inputs and outputs are representing a node or neuron. In addition, there are one or more hidden layers. The objective of the learning phase is to adjust the weights of the connections between different layers or nodes. After setting up the learning samples, in an iterative approach a sample will be fed into the network and

2.1 Artificial Neura ra

l l Network Artifi fi ci cial neural l ne e tw tw ork A A NN NN , u u su u al l ly ly c cal al le led d Neural Netwo work NN, is an a algorithm m th th at was a origina a ll ll y y mo mo ti ti va va te te d d by by the goa o l of f h h av av ing ma ch c ines that can n mi mim mic c the br br ai n. A neural network consists o f an an inter rco conn nne ected gr ou o p of ar ar ti ifi ficial n n eurons. Th ey are phy si cal cellular s ys te ms c ap apable l o o f f obtainin ing, st s orin n g information, and using e xp er iential kn ow ledge. Like hu huma a n br br ai ain n, the e AN N N’s kn ow ledge come s fr om e xamp le s that they en co un te er. In n hu hu man n ne e ur al system, lea rn in g pr oc es s in cl ud es t he mod if ic ations to t the synap ptic c co o nn ection s betwee n the neurons. In a similar way, ANNs a adju u st st the ir ir structur e ba base se d d on on o o ut ut pu pu t t an and inpu pu t t in in fo fo rm rm at at io io n n th th at at flo ws through the he network during the learning phase. e Da Data ta p p ro ro ce ce ss ss in ing g pr pr oc oc ed ed ur ur e e in in a a ny ny typ yp ic ical al n n eu eu ra ra l l ne ne tw tw or or k k ha ha s s tw two o ma majo jor r st steps: th th e le le ar ar ni ni n ng a a nd nd application n step. At t t the first step ep , , a tr tr ai ai ni ni ng ng d d at atabase or historical price data is neede e d to train t t he networks. This dataset includes an input vector and a known ou utput vect o or. Each one of the inputs and outputs are representing a node or neu u ro r n. I I n n addition, there are one or more hidden layers. The objective of the learn rn ing phase is to adjust the weights of the the resulting outputs will be compared with the known outputs. If the result and the unknown output are not equal, changing the weights of the connections will be continued until the difference is minimized. After acquiring the desired convergence for the networks in the learning process, the validation dataset is applied to the network for the validating step Shahkarami A. et al. 2014. Fig. 1 An artificial neural network is an interconnected group of nodes. Source : SPE International, Colorado, USA, 16–18April 2014.

2.2 Review of previous researches