This action might not be possible to undo. Are you sure you want to continue?
SATELLITE COMMUNICATION 2. TALKER IDENTIFICATION CONCLUSION BIBILIOGRAPHY .com CONTENTS INTRODUCTION THE ARTIFICIAL NEURON DESIGN LAYERS COMMUNICATION AND TYPES OF CONNECTION LEARNING AREAS OF APPLICATION 1.101seminartopics.
Thus Neural Networks is an exponentially growing area of real. repair. instruction. Inspired by the computational style of biological systems. In other words. prediction. control. . diagnosis. They are of greatest use in computing problems where the input does not follow clean strict rules but instead has an overall pattern.com Artificial Neural Networks ABSTRACT Most people when asked if they think computers could ever become sentient quickly respond no and refer to the fact that computers are unable to learn. They are processing devices. just like children learn to recognize dogs from examples of dogs and exhibit some structural capability for generalization. Neural Networks encompass a diverse set of computational models.101seminartopics. debugging. The most significant aspects of Neural Networks are that they allow the computer to learn and they have the potential for parallelism. monitoring. a Neural Network can be viewed as an assembly of simple. Neural Networks have applications in diverse areas like interpretation. Neural networks are distinguished from other computer and mathematical techniques by their design motivation. which communicate to each other using unidirectional connections. This means that they allow the computer to solve multiple problems at a time. Neural Networks seems to do just that. interconnected processing units (neurons) acting in parallel. categorization and pattern recognition. Most Neural Networks have some sort of “training” rule whereby the weights of connections are adjusted on the basis of presented patterns. which share a set of simple underlying characteristics. planning. Neural Networks “learn” from examples. Neural Networks can perform any variety of tasks just as any regular computer. However. that can be algorithms or actual hardware that are modeled after the functioning of human brain.time applications of the new era.
the multiple layers of simple processing elements called neurons. machine learning algorithms and artificial neural networks. . Learning is accomplished by adjusting these strengths to cause the overall network to output appropriate results. The field goes by many names.101seminartopics. such as connectionism. It is an attempt to simulate within specialized hardware or sophisticated software. parallel distributed processing. Each neuron is linked to certain of its neighbours with varying coefficients of connectivity that represent the strengths of these connections. However. natural intelligent systems. THE ANALOGY TO BRAIN The most basic components of neural networks are modeled after the structure of the brain.com ARTIFICIAL NEURAL NETWORKS INTRODUCTION Artificial Neural network is a system loosely modeled on the human brain. euro computing. neural networks have a strong similarity to the biological brain and therefore a great deal of the terminology is borrowed from neuroscience. Some neural network structures are not closely to that of the brain and some does not have a biological counterpart in the brain.
Basically. think and apply previous experiences to our every action. This power of the brain comes from the numbers of these basic components and the multiple connections between them. simulates the four basic functions of natural neurons.101seminartopics. and then output the final result. THE ARTIFICIAL NEURON The basic unit of neural networks. performs a generally non-linear operation on the result. . a biological neuron receives inputs from other sources. soma. which are dendrites.com THE BIOLOGICAL NEURON The most basic element of the human brain is a specific type of cell. combines them in some way. All natural neurons have four basic components. which provides us with the abilities to remember. the artificial neurons. axon and synapses. Artificial neurons are much simpler than biological neuron.
In the simplest case. these weights are represented by w (n). these products are simply summed. fed through the transfer function to generate a result and then output. The design issues in neural networks are complex and are the major concerns of system developers.com Note that various inputs to the network are represented by the mathematical symbol. . Each of these inputs are multiplied by a connection weight. Designing a neural network consists of: Arranging neurons in various layers.101seminartopics. the fundamentals may vary in these building blocks and there are differences. DESIGN The developer must go through a period of trial and error in the design decisions before coming up with a satisfactory design. Even though all artificial neural networks are constructed from this basic building block. x (n).
These neurons seem capable of nearly un-restricted interconnections. As the figure below shows.com Deciding the type of connections among neurons for different layers. This clustering occurs by creating layers. the figure below describes its basic steps LAYERS Biologically.101seminartopics. the neurons are grouped into layers. Artificial neural networks are the simple clustering of the primitive artificial neurons. which are then connected to one another. Basically. all artificial neural networks have a similar structure of topology. This is not true in any manmade network. as well as among the neurons within the layer. The output layer consists of neurons that communicate the output of the system to the user or external environment. How these layers connect may also vary. . The process of designing a neural network is an iterative process. There are usually a number of hidden layers between these two layers. neural networks are constructed in a three dimensional way from microscopic components. Determining the strength of connection within the network by allowing the network learns the appropriate values of connection weights by using a training data set. Some of the neurons interface the real world to receive its inputs and other neurons provide the real world with the network’s outputs. The input layer consists of neurons that receive input from the external environment. Deciding the way a neuron receives input and produces output. Al the rest of the neurons are hidden from view.
101seminartopics. one are often left out to the method trial and error. The process continues until a certain condition is satisfied or until the output layer is invoked and fires their output to the external environment. If we increase the hidden number of neurons too much you will get an over fit. that is the net will have problem to generalize. making the network useless on new data set. To determine the number of hidden neurons the network should have to perform its best. which becomes input to the other layers of the system. .com When the input layer receives the input its neurons produce output. The training set of data will be memorized.
INTER-LAYER CONNECTIONS There are different types of connections used between layers are called inter-layer connections.101seminartopics. o Fully connected Each neuron on the first layer is connected to every neuron on the second layer. The neuron in a layer may communicate with each other. but produces a single output. there might however be a twoway connection between two neurons. which is communicated to other neurons. A neuron receives input many neurons. or they may not have any connections. The neurons of one layer are always connected to the neurons of at least another layer.com COMMUNICATION AND TYPES OF CONNECTIONS Neurons are connected via a network of paths carrying the output of one neuron as input to another neuron. o Partially connected A neuron of the first layer does not have to be connected to all neurons on the second layer. . but they do not receive any input back from the neurons on the second layer. o Feed forward The neurons on the first layer send their output to the neurons on the second layer. These paths is normally unidirectional. because there may be another path in reverse direction.
and has inhibitory connections to other neurons. they communicate their outputs with one another a number of times before they are allowed to send their outputs to another layer. o Hierarchical If a neural network has a hierarchical structure. this is known as intra-layer connections. Feed forward and bi-directional functions could be fully or partially connected. One can imagine this type of connection as a competitive gang of neurons. Generally some conditions among the neurons of the layer should be achieved before they communicate their outputs to another layer.com o Bi-directional There is another set of connections carrying the output pf the second layer into the neurons of the first layer. INTRA-LAYER CONNECTIONS In more complex structures the neurons communicate among themselves within a layer. the neurons of a lower layer may only communicate with neurons on the next level of layer.101seminartopics. o Resonance The layers have bi-directional connections. On-center/off surround A neuron within a layer has excitatory connections to itself and its immediate neighbors. After these neurons receive input from another layer. Recurrent The neurons within a layer are fully or partially connected to one another. There are two types of intra-layer connections. Each gang excites it and its neurons of the . and they can continue sending messages across the connections a number of times until a certain condition is achieved.
Neural networks are sometimes called machine-learning algorithms. When the connection type between two neurons is inhibitory.com gang members and inhibits all members of other gangs. One causes the summing mechanism of the next neuron to add while the other causes it to subtract. the neurons with an active output value will win. One excites while the other inhibits. because changing of its connection weights (training) causes the network to learn the solution to a problem. After a few rounds of signal interchange. In the excitatory connection. The learning ability of a neural network is determined by its architecture and by the algorithmic method chosen for training. the output of one neuron increases the action potential of the neuron to which it is connected. The training method usually consists of one or three schemes: 1. excitatory or inhibitory.) LEARNING The brain basically learns from experience. then the output of the neuron sending a message would reduce the activity or action potential of the receiving neuron. In this approach no sample outputs are provided to the network against .101seminartopics. Unsupervised learning The hidden neurons must find a way to organize themselves without help from outside. The system learns new knowledge by adjusting these connection weights. (There are two types of connections between two neurons. The strength of connection between the neurons is stored as a weight-value for the specific connection. and is allowed to update itself and its gang member’s weights.
because it requires a teacher. Both unsupervised and reinforcement suffers from relative slowness and inefficiency relying on a random shuffling to find the proper connection weights. This is also sometimes called recall. thus improving performance. The connections among the neurons in the hidden layer are randomly arranged. When the system is being used as a decision aid to make recommendations. A form of supervised learning. This is learning by doing. The teacher may be a training set of data or an observer who grades the performance of the network results. it is in the operation mode. the system could be in training mode or learning mode. then reshuffled as the network is told how close it is to solving the problem. Back propagation This method is proven highly successful in training of multi-layered neural nets. 2. Reinforcement learning This method works on reinforcement from outside.com which it can measure its predictive performance for a given vector of inputs. Reinforcement learning is also called supervised learning. The network is not just given reinforcement for how it is doing on a task. 3.101seminartopics. . When the system uses input data to change its weights to learn the domain knowledge. Information about errors is also filtered back through the system and is used to adjust the connections between the layers. OFF-LINE OR ON-LINE One can categorize the learning methods into yet another group. off-line or on-line.
E. On-line In on-line or real-time learning. E. This type of learning has a more complex design structure. .com Off-line In the off-line learning methods. once the system enters into operation mode.g. a pattern. AREAS OF APPLICATION Basically most applications of neural networks fall into the following five categories: Prediction: Uses input values to predict some output. a picture. it continues to learn while being used as a decision tool. Data-conceptualization: Analyze the inputs so that grouping relationships can be inferred. a chemical compound. Classification of the object under consideration as any one of the categories (An illness. its weights are fixed and do not change any more. E.g.g.) That in return may trigger the recommendation of an action. a word etc. identifies when the scanner is not working properly. Classification: Uses input values to determine the classification. Most of the networks are of the off-line learning type. Pick the best stocks in the market. Data association: It recognizes the data that contains errors. when the system is in operating mode (recall).101seminartopics. predict whether and identify patients with cancer risk.
the names of those most likely to buy a particular product. .101seminartopics. takes the noise out of a telephone signal. Two schemes are considered in this approach THE SELF-ORGANISATION METHOD The scheme consists of two levels of management as shown in the figure: First-level management -Map configuration by self-organization. E.com E. thus improving the grade of service while maximizing network utilization.g. Data filtering: Smoothes an input signal. The approach allows the network to change according to the input pattern. which incorporates the idea of dynamically adapting the network as well as dynamically routing each arrival.g. NEURAL NETWORKS IN SATELLITE COMMUNICATION Here an approach is introduced for efficient traffic management for a satellite network of a geo-stationary orbital type. Extract from a data base. Second Level Management -Routing Algorithm The first level adaptively configures maps for the satellite communication network and the second level routes traffic under the fixed configuration.
which is continuously updated by a better one. Further given a configuration. we perform the routing dynamically. In text-independent processing the talker is identified by constructing an average acoustic model of the talker’s speech independent of the specific words spoken and vice-versa. Map Generator Arbitrator Router Controller NEURAL NETWORK IN TALKER IDENTIFICATION There are two types of talker identification systems text-independent and text-dependent.com THE COST MINIMISATION METHOD In this scheme we will find a configuration that will result in an optimal network wide performance.101seminartopics. BLOCK-DIAGRAM OF THE TALKER-IDENTIFICATION SYSTEM Training Short-time spectra Clustering Talker-models -------------------------------------------------------------------------------------identification Short-time spectra Distance-measures decision . The figure for traffic management scheme using the cost-minimization method.
BIBILIOGRAPHY NEURAL NETWORKS AND TELECOMMUNICATIONS By Ben Yuhas and Nerwan Ansari. . In identification phase. which reduce the number of parameters.com The diagram is divided into two phases: Training and Identification phases. talker models are constructed from short time spectra of the speed signal. The spectra are encoded into cluster sets. In addition the ability of neural networks to learn allows them to be used on problems where straightforward heuristic or rule-based solutions donot exist. A separate cluster set is formed for each talker. The model which generates the least score (distance) is identified as the talker. CONCLUSION Neural networks offer the computational power of non-linear technique while providing a natural path to efficient massively parallel hardware implementations. In training. short time spectra are again computed. Together these capabilities allow Neural Networks to offer unique solutions to problems in telecommunications and other fields.101seminartopics. and a score is generated for each. They are compared with a set of N talker models. NEURAL NETWORKS By Haykin Simon ARTIFICIAL NEURAL NETWORK TECHNOLOGY Data and analysis center for software.
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue listening from where you left off, or restart the preview.