Spelling suggestions: "subject:"artificial aeural"" "subject:"artificial aneural""
1 |
Neural models of temporal sequencesTaylor, Neill Richard January 1998 (has links)
No description available.
|
2 |
Neural implementations of canonical correlation analysisLai, Pei Ling January 2000 (has links)
No description available.
|
3 |
Condition monitoring of rotating machinery using wavelets as pre-processor to ANNsEngin, Seref Naci January 1998 (has links)
No description available.
|
4 |
Image Inpainting Based on Artifical Neural NetworksHsu, Chih-Ting 29 June 2007 (has links)
Application of Image inpainting ranges from object removal, photo restoration, scratch removal, and so on. In this thesis, we will propose a modified multi-scale method and learning-based method using artificial neural networks for image inpainting.
Multi-scale inpainting method combines image segmentation, contour estimation, and exemplar-based inpainting. The main goal of image segmentation is to separate image to several homogeneous regions outside the target region. After image segmentation, we use contour estimation to estimate curves inside the target region to partition the whole image into several different regions. Then we fill those different regions inside the target region separately by exemplar-based inpainting method.
The exemplar-based technique fills the target region via the texture synthesis and filling order of exemplary patches. Exemplary patches are found near target region and the filling order is determined by isophote and densities of exemplary patches.
Learning-based inpainting is a novel technique. This technique combines machine learning and the concept of filling order. We use artificial neural networks
to learn the structure and texture surrounding the target region. After training, we fill the target region according to the filling order.
From our simulation results, very good results can be obtained for removing large-size objects by using the proposed multi-scale method, and for removing medium-size objects of gray images.
|
5 |
Evolving Turing's Artificial Neural NetworksOrr, Ewan January 2010 (has links)
Our project uses ideas first presented by Alan Turing. Turing's immense contribution to mathematics and computer science is widely known, but his pioneering work in artificial intelligence is relatively unknown. In the late 1940s Turing introduced discrete Boolean artificial neural networks and, it has been argued that, he suggested that these networks be trained via evolutionary algorithms. Both artificial neural networks and evolutionary algorithms are active fields of research. Turing's networks are very basic yet capable of complex tasks such as processing sequential input; consequently, they are an excellent model for investigating the application of evolutionary algorithms to artificial neural networks. We define an example of these networks using sequential input and output, and we devise evolutionary algorithms that train these networks. Our networks are discrete Boolean networks where every 'neuron' either performs NAND or identity, and they can represent any function that maps one sequence of bit strings to another. Our algorithms use supervised learning to discover networks that represent such functions. That is, when searching for a network that represents a particular function our algorithms use input-output pairs of that function as examples to aid the discovery of solution networks. To test our ideas we encode our networks and implement the algorithms in a computer program. Using this program we investigate the performance of our networks and algorithms on simple problems such as searching for networks that realize the parity function and the multiplexer function. This investigation includes the construction and testing of an intricate crossover operator. Because our networks are composed of simple 'neurons' they are a suitable test-bed for novel training schemes. To improve our evolutionary algorithms for some problems we employ the symmetry of the problem to reduce its search space. We devise and test a means of using subgroups of the group of permutation of inputs of a function to aid evolutionary searches search for networks that represent that function. In particular, we employ the action of the permutation group S₂ to 'cut down' the search space when we search for networks that represent functions such as parity.
|
6 |
A comparison of encoding schemes for neural network evolutionSiddiqi, Abdul Ahad January 1998 (has links)
No description available.
|
7 |
An Artificial neural network-based signal classifier for automated identification of detection signals from a dielectrophoretic cytometerBhide, Ashlesha 26 February 2014 (has links)
An automated signal classifier and a semi-automated signal identifier are designed for collecting the dielectrophoretic signatures of cells flowing through a dielectrophoretic cytometer. In past work, the DEP cytometer signals were manually sorted by going through all recorded signals, which is impractical when analyzing 1000’s of cells per day. In the semi-automated method of collection, signals are automatically identified as events and displayed on the user interface to be accepted or rejected by the user. This approach reduced signal collection time by more than half and produced statistics nearly identical to the manual method. The automated signal classifier based on pattern recognition categorizes detection signals as ‘Accept’ or ‘Reject’. Analyzing large volumes of detection signals is possible in much reduced times and may be approaching real time capability.
|
8 |
The application and analysis of genetic algorithms to discover topological free parameters in multi-layer perceptionsKrasniewicz, Jan A. January 2000 (has links)
No description available.
|
9 |
An Artificial neural network-based signal classifier for automated identification of detection signals from a dielectrophoretic cytometerBhide, Ashlesha 26 February 2014 (has links)
An automated signal classifier and a semi-automated signal identifier are designed for collecting the dielectrophoretic signatures of cells flowing through a dielectrophoretic cytometer. In past work, the DEP cytometer signals were manually sorted by going through all recorded signals, which is impractical when analyzing 1000’s of cells per day. In the semi-automated method of collection, signals are automatically identified as events and displayed on the user interface to be accepted or rejected by the user. This approach reduced signal collection time by more than half and produced statistics nearly identical to the manual method. The automated signal classifier based on pattern recognition categorizes detection signals as ‘Accept’ or ‘Reject’. Analyzing large volumes of detection signals is possible in much reduced times and may be approaching real time capability.
|
10 |
Design and application of neurocomputersNaylor, David C. J. January 1994 (has links)
This thesis aims to understand how to design high performance, flexible and cost effective neural computing systems and apply them to a variety of real-time applications. Systems of this type already exist for the support of a range of ANN models. However, many of these designs have concentrated on optimising the architecture of the neural processor and have generally neglected other important aspects. If these neural systems are to be of practical benefit to researchers and allow complex neural problems to be solved efficiently, all aspects of their design must be addressed.
|
Page generated in 0.0496 seconds