Parallel path consistency

Filtering algorithms are well accepted as a means of speeding up the solution of the consistent labeling problem (CLP). Despite the fact that path consistency does a better job of filtering than arc consistency, AC is still the preferred technique because it has a much lower time complexity. We are implementing parallel path consistency algorithms on multiprocessors and comparing their performance to the best sequential and parallel arc consistency algorithms.^(1,2) (See also work by Kerethoet al. ⁽³⁾ and Kasif⁽⁴⁾) Preliminary work has shown linear performance increases for parallelized path consistency and also shown that in many cases performance is significantly better than the theoretical worst case. These two results lead us to believe that parallel path consistency may be a superior filtering technique. Finally, we have implemented path consistency as an outer product computation and have obtained good results (e.g., linear speedup on a 64K-node Connection Machine 2).     

A prototype 3D optically interconnected neural network

We report the implementation of a prototype three-dimensional (3D) optoelectronic neural network that combines free-space optical interconnects with silicon-VLSI-based optoelectronic circuits. The prototype system consists of a 16-node input, 4-neuron hidden, and a single-neuron output layer, where the denser input-to-hidden-layer connections are optical. The input layer uses PLZT light modulators to generate optical outputs which are distributed over an optoelectronic neural network chip through space-invariant holographic optical interconnects. Optical interconnections provide negligible fan-out delay and allow compact, purely on-chip electronic H-tree type fan-in structure. The small prototype system achieves a measured 8-bit electronic fan-in precision and a calculated maximum speed of 640 million interconnections per second. The system was tested using synaptic weights learned off system and was shown to distinguish any vertical line from any horizontal one in an image of 4×4 pixels. New, more efficient light detector and small-area analog synapse circuits and denser optoelectronic neuron layouts are proposed to scale up the system. A high-speed, feed-forward optoelectronic synapse implementation density of up to 10⁴/cm² seems feasible using new synapse design. A scaling analysis of the system shows that the optically interconnected neural network implementation can provide higher fan-in speed and lower power consumption characteristics than a purely electronic, crossbar-based neural network implementation     

The development of a stress analysis code for nuclear graphite components in gas-cooled reactors

Most of the UK nuclear power reactors are gas-cooled and graphite moderated. As well as acting as a moderator the graphite also acts as a structural component providing channels for the coolant gas and control rods. For this reason the structural integrity assessments of nuclear graphite components is an essential element of reactor design. In order to perform graphite component stress analysis, the definition of the constitutive equation relating stress and strain for irradiated graphite is required. Apart from the usual elastic and thermal strains, irradiated graphite components are subject to additional strains due to fast neutron irradiation and radiolytic oxidation. In this paper a material model for nuclear graphite is presented along with an example of a stress analysis of a nuclear graphite moderator brick subject to both fast neutron irradiation and radiolytic oxidation.     

The Effects of Cyclic Terminal Groups in 4-Aminoazo-benzene and Related Azo Dyes: Part 21—pKa Values of Some Monoazo Dyes Derived from

Basicity values have been determined for series of monoazo dyes incorporating a pyrrolidino or a piperidino moiety. In general, the pK_a value is found to decrease as the donor capacity of the terminal nitrogen atom decreases.The basicities of the 4'-substituted piperidinoazobenzenes decrease as the electron withdrawing strength of the 4'-substituent increases. However, the pK_a values of the related pyrrolidino compounds appear to be insensitive to the substituent.     

Characterization of platelet-activating factor synthesis in glomerular endothelial cell lines

Platelet-activating factor synthesis in two transformed lines of glomerular endothelial cells was characterized and contrasted with platelet-activating factor production in macrovascular-derived endothelial cells as well as with glomerular cells of mesenchymal origin. Platelet-activating factor synthesis was assessed in intact cells and in cell-free preparations. Glomerular endothelial cells constitutively synthesize bio-active alkyl-PAF, and this basal activity can be chronically augmented by various inflammatory and thrombotic agents. In contrast, thrombin-mediated platelet-activating factor formation in bovine pulmonary aortic endothelial cells as well as in glomerular mesangial cells is acute and transient. The potential role of anti-inflammatory prostanoids to function as negative feedback modulators of thrombin- or endothelin-mediated platelet-activating factor synthesis was also investigated, as the synthesis of platelet-activating factor is often associated with the formation of these prostanoids. Indomethacin augmented receptor-mediated platelet-activating factor synthesis while prostanoids of the E and I series reduced agonist-stimulated PAF synthesis. In summary, the unique capacity of glomerular endothelial cells to respond to inflammatory stimuli with sustained platelet-activating factor synthesis is a clear indication of this cell's pivotal role in augmenting the inflammatory response in the limited environment of the glomerulus.     

Modulation of postural tremors at the wrist by supramaximal electrical median nerve shocks in essential tremor, Parkinson's disease and normal subjects mimicking tremor

The response of postural wrist tremors to supramaximal median nerve stimulation was examined in patients with hereditary essential tremor (n = 10) and Parkinson's disease (n = 9), and in normal subjects mimicking wrist tremor (n = 8). The average frequency of on-going tremor was the same in all three groups. Supramaximal peripheral nerve shocks inhibited and then synchronised the rhythmic electromyographic (EMG) activity of all types of tremor. The duration of inhibition ranged from 90 to 210ms, varying inversely with the frequency of on-going tremor. There was no significant difference in mean duration of inhibition or in the timing of the first peak after stimulation on the average rectified EMG records between the three groups. The degree to which supramaximal peripheral nerve shocks could modulate the timing of rhythmic EMG bursts in the forearm flexor muscles was also quantified by deriving a resetting index. No significant difference in mean resetting index of the three groups was found. These results suggest that such studies cannot be used to differentiate between the common causes of postural wrist tremors.     

Use of climate scenarios to aid in decision analysis for interannual water supply planning

This work addresses the issue of climate change in the context of water resource planning on the time scale of a few years. Planning on this time scale generally ignores the role of climate change. However, where the climate of a region has already shifted, the use of historical data for planning purposes may be misleading. In order to test this, a case study is conducted for a region, the Australian Capital Territory, where long term drought is raising concerns of a possible climate shift. The issue is cast in terms of a particular planning decision; the option to augment water supply in the next few years to hedge against the drought persisting. A set of climate scenarios are constructed for the region corresponding to the historical climate regime and to regimes where progressively greater levels of change are assumed to have already taken place (5%, 10%, 20% reductions in mean rainfall). Probabilities of the drought persisting are calculated for each of the scenarios. The results show substantial increases in the probability of the drought persisting for even moderate reductions in mean rainfall. The sensitivity of the decision to augment supply to the scenario results depends ultimately on the planners tolerable thresholds for the probability of the drought persisting. The use of different scenarios enables planners to explore the sensitivity of the decision in terms of their risk tolerance to ongoing drought and to their degree of belief in each of the scenarios tested.     

Rainbow smelt population responses to species invasions and change in environmental condition

Invasive species can cause major disruptions in native food webs, yet the impact of species introductions and whether they will become invasive appears to be context-dependent. Rainbow smelt and alewife co-exist as invasive species in the Laurentian Great Lakes and as native species on the Atlantic coast of North America, but in Lake Champlain rainbow smelt is the dominant native forage fish and alewife are invasive. Alewife became abundant by 2007, providing an opportunity to explore the dynamics of these two species in a system where only one is invasive. We used data from a 28-year forage fish survey to compare demographics of rainbow smelt populations in three basins of Lake Champlain with different volumes, nutrient levels, and predator abundances. Rainbow smelt catch-per-unit-effort (CPUE) remained constant in the large, deep Main Lake before and after alewife invaded, but decreased in the two smaller basins. Declines were primarily a result of increased age-0 and age-1 mortality. Predation by top piscivores, system productivity, and competition for resources alone could not explain the patterns in CPUE across the basins. The mechanisms that allow alewife and rainbow smelt to co-exist could be related to system volume and oxythermal habitat availability, and may explain why the two species do not negatively affect each other in the Great Lakes. Summer hypoxia in the smaller basins could force individuals into smaller habitat volumes with higher densities of competitors and cannibalistic adult rainbow smelt. Habitat availability may mediate the impact of invasive alewife on native rainbow smelt.     

Generation of optimal artificial neural networks using a pattern search algorithm: application to approximation of chemical systems

A pattern search optimization method is applied to the generation of optimal artificial neural networks (ANNs). Optimization is performed using a mixed variable extension to the generalized pattern search method. This method offers the advantage that categorical variables, such as neural transfer functions and nodal connectivities, can be used as parameters in optimization. When used together with a surrogate, the resulting algorithm is highly efficient for expensive objective functions. Results demonstrate the effectiveness of this method in optimizing an ANN for the number of neurons, the type of transfer function, and the connectivity among neurons. The optimization method is applied to a chemistry approximation of practical relevance. In this application, temperature and a chemical source term are approximated as functions of two independent parameters using optimal ANNs. Comparison of the performance of optimal ANNs with conventional tabulation methods demonstrates equivalent accuracy by considerable savings in memory storage. The architecture of the optimal ANN for the approximation of the chemical source term consists of a fully connected feedforward network having four nonlinear hidden layers and 117 synaptic weights. An equivalent representation of the chemical source term using tabulation techniques would require a 500 x 500 grid point discretization of the parameter space.