No description available.

The Where's Waldo problem concerns how individuals can rapidly scan a scene to detect a target object in it. This dissertation develops the ARTSCAN Search neural model to clarify how brain mechanisms that govern spatial and object attention, spatially-invariant object learning and recognition, reinforcement learning, and eye movement search are coordinated to enable learning and directed search for desired objects at specific locations in a cluttered scene. In the model, interactions from the Where cortical processing stream to the What cortical processing stream modulate invariant category learning of a desired object, whereas interactions from the What cortical processing stream to the Where cortical processing stream support search for the object. In particular, when an invariant object category representation is activated top-down by a cognitive plan or by an active motivational source in the model's What stream, it can shift spatial attention in the Where stream and thereby selectively activate the locations of sought-after object exemplars. These combined What-to-Where and Where-to-What interactions clarify how the brain's solution of the Where's Waldo problem overcomes the complementary deficiencies of What and Where stream processes taken individually by using inter-stream interactions that allow both invariant object recognition and spatially selective attention and action to occur.

This dissertation introduces advanced artificial intelligence based algorithm for detecting and classifying faults on the power system transmission line. The proposed algorithm is aimed at substituting classical relays susceptible to possible performance deterioration during variable power system operating and fault conditions. The new concept relies on a principle of pattern recognition and detects the existence of the fault, identifies fault type, and estimates the transmission line faulted section. The approach utilizes self-organized, Adaptive Resonance Theory (ART) neural network, combined with fuzzy decision rule for interpretation of neural network outputs. Neural network learns the mapping between inputs and desired outputs through processing a set of example cases. Training of the neural network is based on the combined use of unsupervised and supervised learning methods. During training, a set of input events is transformed into a set of prototypes of typical input events. During application, real events are classified based on the interpretation of their matching to the prototypes through fuzzy decision rule. This study introduces several enhancements to the original version of the ART algorithm: suitable preprocessing of neural network inputs, improvement in the concept of supervised learning, fuzzyfication of neural network outputs, and utilization of on-line learning. A selected model of an actual power network is used to simulate extensive sets of scenarios covering a variety of power system operating conditions as well as fault and disturbance events. Simulation results show improved recognition capabilities compared to a previous version of ART neural network algorithm, Multilayer Perceptron (MLP) neural network algorithm, and impedance based distance relay. Simulation results also show exceptional robustness of the novel ART algorithm for all operating conditions and events studied, as well as superior classification capabilities compared to the other solutions. Consequently, it is demonstrated that the proposed ART solution may be used for accurate, high-speed distinction among faulted and unfaulted events, and estimation of fault type and fault section.

Most of virtual communities are built on the client/server system. There are some limitations on the client/server system such as the maintenance cost and the personal attribute protection. The peer-to-peer system has some strengths to overcome the limitations of client/server system. Therefore, we are willing to export the virtual community on the peer-to-peer system. There are two main team formation approaches in the current virtual community collaboration. Either one of these approaches alone has its limitations. In this study, we adopt the social network concept to design a team formation mechanism in order to overcome the limitations of current approaches. Besides, because of the natural of peer-to-peer system, the exchange of messages is sending and receiving on the network. The mechanism proposed in this research can also reduce the traffic cost of the team formation process. Furthermore, it maintains the fitness of members who are chosen in the same team.

This thesis introduces a novel behavior categorization model that can be used for behavior recognition and learning. Correlation Based Adaptive Resonance Theory (CobART) network, which is a kind of self organizing and unsupervised competitive neural network, is developed for this purpose. CobART uses correlation analysis methods for category matching. It has modular and simple architecture. It can be adapted to different categorization tasks by changing the correlation analysis methods used when needed. CobART networks are integrated hierarchically for an adequate categorization of behaviors. The hierarchical model is developed by adding a second layer CobART network on top of first layer networks. The first layer CobART networks categorize self behavior data of a robot or an object in the environment. The second layer CobART network receives first layer CobART network categories as an input, and categorizes them to elicit the robot&#039 / s behavior with respect to its effect on the object. Besides, the second layer network back-propagates the matching information to the first layer networks in order to find the relation between the first layer categories. The performance of the hierarchical model is compared with that of different neural network based models. Experiments show that the proposed model generates reasonable categorization of behaviors being tested. Moreover, it can learn different forms of the behaviors, and it can detect the relations between them. In essence, the model has an expandable architecture and it contains reusable parts. The first layer CobART networks can be integrated with other CobART networks for another categorization task. Hence, the model presents a way to reveal all behaviors performed by the robot at the same time.

This dissertation introduces advanced artificial intelligence based algorithm for detecting and classifying faults on the power system transmission line. The proposed algorithm is aimed at substituting classical relays susceptible to possible performance deterioration during variable power system operating and fault conditions. The new concept relies on a principle of pattern recognition and detects the existence of the fault, identifies fault type, and estimates the transmission line faulted section. The approach utilizes self-organized, Adaptive Resonance Theory (ART) neural network, combined with fuzzy decision rule for interpretation of neural network outputs. Neural network learns the mapping between inputs and desired outputs through processing a set of example cases. Training of the neural network is based on the combined use of unsupervised and supervised learning methods. During training, a set of input events is transformed into a set of prototypes of typical input events. During application, real events are classified based on the interpretation of their matching to the prototypes through fuzzy decision rule. This study introduces several enhancements to the original version of the ART algorithm: suitable preprocessing of neural network inputs, improvement in the concept of supervised learning, fuzzyfication of neural network outputs, and utilization of on-line learning. A selected model of an actual power network is used to simulate extensive sets of scenarios covering a variety of power system operating conditions as well as fault and disturbance events. Simulation results show improved recognition capabilities compared to a previous version of ART neural network algorithm, Multilayer Perceptron (MLP) neural network algorithm, and impedance based distance relay. Simulation results also show exceptional robustness of the novel ART algorithm for all operating conditions and events studied, as well as superior classification capabilities compared to the other solutions. Consequently, it is demonstrated that the proposed ART solution may be used for accurate, high-speed distinction among faulted and unfaulted events, and estimation of fault type and fault section.

No description available.

This thesis focuses on simulating, detecting, localizing and classifying the power quality disturbances using advanced signal processing techniques and neural networks. Primarily discrete wavelet and Fourier transforms are used for feature extraction, and classification is achieved by using neural network algorithms. The proposed feature vector consists of a combination of features computed using multi resolution analysis and discrete Fourier transform. The proposed feature vectors exploit the benefits of having both time and frequency domain information simultaneously. Two different classification algorithms based on Feed forward neural network and adaptive resonance theory neural networks are proposed for classification. This thesis demonstrates that the proposed methodology achieves a good computational and error classification efficiency rate.

This thesis aims at detecting and classifying the power system transmission line faults. To deal with the problem of an extremely large data set with different fault situations, a three step optimized Neural Network approach has been proposed. The approach utilizes Discrete Wavelet Transform for detection and two different types of self-organized, unsupervised Adaptive Resonance Theory Neural Networks for classification. The fault scenarios are simulated using Alternate Transients Program and the performance of this highly improved scheme is compared with the existing techniques. The simulation results prove that the proposed technique handles large data more efficiently and time of operation is considerably less when compared to the existing methods.

This work presents a proof of concept of an Unsupervised Learning Trojan. The Unsupervised Learning Trojan presents new challenges over previous work on the Neural network Trojan, since the attacker does not control most of the environment. The current work will presented an analysis of how the attack can be successful by proposing new assumptions under which the attack can become a viable one. A general analysis of how the compromise can be theoretically supported is presented, providing enough background for practical implementation development. The analysis was carried out using 3 selected algorithms that can cover a wide variety of circumstances of unsupervised learning. A selection of 4 encoding schemes on 4 datasets were chosen to represent actual scenarios under which the Trojan compromise might be targeted. A detailed procedure is presented to demonstrate the attack's viability under assumed circumstances. Two tests of hypothesis concerning the experimental setup were carried out which yielded acceptance of the null hypothesis. Further discussion is contemplated on various aspects of actual implementation issues and real world scenarios where this attack might be contemplated.