1 |
Cleanup Memory in Biologically Plausible Neural NetworksSingh, Raymon January 2005 (has links)
During the past decade, a new class of knowledge representation has emerged known as structured distributed representation (SDR). A number of schemes for encoding and manipulating such representations have been developed; e. g. Pollack's Recursive Auto-Associative Memory (RAAM), Kanerva's Binary Spatter Code (BSC), Gayler's MAP encoding, and Plate's Holographically Reduced Representations (HRR). All such schemes encode structural information throughout the elements of high dimensional vectors, and are manipulated with rudimentary algebraic operations. <br /><br /> Most SDRs are very compact; components and compositions of components are all represented as fixed-width vectors. However, such compact compositions are unavoidably noisy. As a result, resolving constituent components requires a cleanup memory. In its simplest form, cleanup is performed with a list of vectors that are sequentially compared using a similarity metric. The closest match is deemed the cleaned codevector. <br /><br /> While SDR schemes were originally designed to perform cognitive tasks, none of them have been demonstrated in a neurobiologically plausible substrate. Potentially, mathematically proven properties of these systems may not be neurally realistic. Using Eliasmith and Anderson's (2003) Neural Engineering Framework, I construct various spiking neural networks to simulate a general cleanup memory that is suitable for many schemes. <br /><br /> Importantly, previous work has not taken advantage of parallelization or the high-dimensional properties of neural networks. Nor have they considered the effect of noise within these systems. As well, additional improvements to the cleanup operation may be possible by more efficiently structuring the memory itself. In this thesis I address these lacuna, provide an analysis of systems accuracy, capacity, scalability, and robustness to noise, and explore ways to improve the search efficiency.
|
2 |
Cleanup Memory in Biologically Plausible Neural NetworksSingh, Raymon January 2005 (has links)
During the past decade, a new class of knowledge representation has emerged known as structured distributed representation (SDR). A number of schemes for encoding and manipulating such representations have been developed; e. g. Pollack's Recursive Auto-Associative Memory (RAAM), Kanerva's Binary Spatter Code (BSC), Gayler's MAP encoding, and Plate's Holographically Reduced Representations (HRR). All such schemes encode structural information throughout the elements of high dimensional vectors, and are manipulated with rudimentary algebraic operations. <br /><br /> Most SDRs are very compact; components and compositions of components are all represented as fixed-width vectors. However, such compact compositions are unavoidably noisy. As a result, resolving constituent components requires a cleanup memory. In its simplest form, cleanup is performed with a list of vectors that are sequentially compared using a similarity metric. The closest match is deemed the cleaned codevector. <br /><br /> While SDR schemes were originally designed to perform cognitive tasks, none of them have been demonstrated in a neurobiologically plausible substrate. Potentially, mathematically proven properties of these systems may not be neurally realistic. Using Eliasmith and Anderson's (2003) Neural Engineering Framework, I construct various spiking neural networks to simulate a general cleanup memory that is suitable for many schemes. <br /><br /> Importantly, previous work has not taken advantage of parallelization or the high-dimensional properties of neural networks. Nor have they considered the effect of noise within these systems. As well, additional improvements to the cleanup operation may be possible by more efficiently structuring the memory itself. In this thesis I address these lacuna, provide an analysis of systems accuracy, capacity, scalability, and robustness to noise, and explore ways to improve the search efficiency.
|
3 |
Forecasting Parameter of Kailashtilla Gas Processing Plant Using Neural NetworkKundu, S., Hasan, A., Sowgath, Md Tanvir 22 December 2012 (has links)
No / Neural Network (NN) is widely used in all aspects of
process engineering activities, such as modeling, design,
optimization and control. In this paper work, in absence of real
plant data, simulated data (such as sales gas flow rate, pressure,
raw gases flow rates and input heat flow associated with a heater
used after dehydration) from a detailed model of Kailashtilla gas
processing plant (KGP) within HYSYS is used to develop NN
based model. Thereafter NN based model is trained and
validated from HYSYS simulator generated data and that
framework can predict the output data (sales gas flow rate and
pressure) very closely with the simulated HYSYS plant data.
The preliminary results show that the NN based correlation is
adequately able to model and generate workable profiles for the
process.
|
4 |
Short term energy forecasting techniques for virtual power plantsRavichandran, S., Vijayalakshmi, A., Swarup, K.S., Rajamani, Haile S., Pillai, Prashant 06 October 2016 (has links)
Yes / The advent of smart meter technology has enabled periodic monitoring of consumer energy consumption. Hence, short term energy forecasting is gaining more importance than conventional load forecasting. An Accurate forecasting of energy consumption is indispensable for the proper functioning of a virtual power plant (VPP). This paper focuses on short term energy forecasting in a VPP. The factors that influence energy forecasting in a VPP are identified and an artificial neural network based energy forecasting model is built. The model is tested on Sydney/ New South Wales (NSW) electricity grid. It considers the historical weather data and holidays in Sydney/ NSW and forecasts the energy consumption pattern with sufficient accuracy.
|
Page generated in 0.0704 seconds