Forgetting Exceptions is Harmful in Language Learning

We show that in language learning, contrary to received wisdom, keeping exceptional training instances in memory can be beneficial for generalization accuracy. We investigate this phenomenon empirically on a selection of benchmark natural language processing tasks: grapheme-to-phoneme conversion, part-of-speech tagging, prepositional-phrase attachment, and base noun phrase chunking. In a first series of experiments we combine memory-based learning with training set editing techniques, in which instances are edited based on their typicality and class prediction strength. Results show that editing exceptional instances (with low typicality or low class prediction strength) tends to harm generalization accuracy. In a second series of experiments we compare memory-based learning and decision-tree learning methods on the same selection of tasks, and find that decision-tree learning often performs worse than memory-based learning. Moreover, the decrease in performance can be linked to the degree of abstraction from exceptions (i.e., pruning or eagerness). We provide explanations for both results in terms of the properties of the natural language processing tasks and the learning algorithms.     

Analytical solution of stresses and material birefringence inoptical fibers with noncircular cladding

An analytical method for the calculation of thermal stress in polarization-maintaining optical fibers is presented. The method is based on the thermoelastic potential and is appropriate for structures with a noncircular stress-cladding boundary. The analytical solution is compared with the results of the finite element method (FEM) for the case of an elliptical boundary     

Signalling and obfuscation for congestion control

We aim to reduce the social cost of congestion in many smart city applications. In our model of congestion, agents interact over limited resources after receiving signals from a central agent that observes the state of congestion in real time. Under natural models of agent populations, we develop new signalling schemes and show that by introducing a non-trivial amount of uncertainty in the signals, we reduce the social cost of congestion, i.e., improve social welfare. The signalling schemes are efficient in terms of both communication and computation, and are consistent with past observations of the congestion. Moreover, the resulting population dynamics converge under reasonable assumptions.     

Zapotin,a Polymethoxyflavone,with Potential Therapeutic Attributes

The use of plants as traditional medicines is common and has prevailed in many different cultures over time. Polymethoxyflavones (PMFs) are natural polyphenols from the group of flavonoids. Zapotin, a member of the PMFs, is found mainly in citrus plants and is almost exclusively limited to their peels. The chemical structure of zapotin has been questioned from the very beginning, since the structure of flavonoids with a single oxygen atom in the C2′ position is extremely rare in the plant kingdom. To clarify this, the structural determination and bio-inspired synthesis of zapotin are discussed in detail in this review. Due to the broad biological potential of PMFs, the complication in the isolation process and characterization of PMFs, as well as their purification, have been estimated by adapting various chromatographic methods. According to available data from the literature, zapotin may be a promising curative agent with extensive biological activities, especially as a chemopreventive factor. Apart from that, zapotin acts as an antidepressant-like, anticancer, antifungal, and antioxidant agent. Finally, accessible studies about zapotin metabolism (absorption, distribution, metabolism, excretion, and toxicity) underline its potential in use as a therapeutic substance.     

Anti-Inflammatory Effect of Very High Dose Local Vessel Wall Statin Administration: Poly(L,L-Lactide) Biodegradable Microspheres with Simvastatin for Drug Delivery System (DDS)

Atherosclerosis involves an ongoing inflammatory response of the vascular endothelium and vessel wall of the aorta and vein. The pleiotropic effects of statins have been well described in many in vitro and in vivo studies, but these effects are difficult to achieve in clinical practice due to the low bioavailability of statins and their first-pass metabolism in the liver. The aim of this study was to test a vessel wall local drug delivery system (DDS) using PLA microstructures loaded with simvastatin. Wistar rats were fed high cholesterol chow as a model. The rat vessels were chemically injured by repeated injections of perivascular paclitaxel and 5-fluorouracil. The vessels were then cultured and treated by the injection of several concentrations of poly(L,L-lactide) microparticles loaded with the high local HMG-CoA inhibitor simvastatin (0.58 mg/kg) concentration (SVPLA). Histopathological examinations of the harvested vessels and vital organs after 24 h, 7 days and 4 weeks were performed. Microcirculation in mice as an additional test was performed to demonstrate the safety of this approach. A single dose of SVPLA microspheres with an average diameter of 6.4 μm and a drug concentration equal to 8.1% of particles limited the inflammatory reaction of the endothelium and vessel wall and had no influence on microcirculation in vivo or in vitro. A potent pleiotropic (anti-inflammatory) effect of simvastatin after local SVPLA administration was observed. Moreover, significant concentrations of free simvastatin were observed in the vessel wall (compared to the maximum serum level). In addition, it appeared that simvastatin, once locally administered as SVPLA particles, exerted potent pleiotropic effects on chemically injured vessels and presented anti-inflammatory action. Presumably, this effect was due to the high local concentrations of simvastatin. No local or systemic side effects were observed. This approach could be useful for local simvastatin DDSs when high, local drug concentrations are difficult to obtain, or systemic side effects are present.     

G-Quadruplex in Gene Encoding Large Subunit of Plant RNA Polymerase II: A Billion-Year-Old Story

G-quadruplexes have long been perceived as rare and physiologically unimportant nucleic acid structures. However, several studies have revealed their importance in molecular processes, suggesting their possible role in replication and gene expression regulation. Pathways involving G-quadruplexes are intensively studied, especially in the context of human diseases, while their involvement in gene expression regulation in plants remains largely unexplored. Here, we conducted a bioinformatic study and performed a complex circular dichroism measurement to identify a stable G-quadruplex in the gene RPB1, coding for the RNA polymerase II large subunit. We found that this G-quadruplex-forming locus is highly evolutionarily conserved amongst plants sensu lato (Archaeplastida) that share a common ancestor more than one billion years old. Finally, we discussed a new hypothesis regarding G-quadruplexes interacting with UV light in plants to potentially form an additional layer of the regulatory network.     

Accelerating scientific computations with mixed precision algorithms

On modern architectures, the performance of 32-bit operations is often at least twice as fast as the performance of 64-bit operations. By using a combination of 32-bit and 64-bit floating point arithmetic, the performance of many dense and sparse linear algebra algorithms can be significantly enhanced while maintaining the 64-bit accuracy of the resulting solution. The approach presented here can apply not only to conventional processors but also to other technologies such as Field Programmable Gate Arrays (FPGA), Graphical Processing Units (GPU), and the STI Cell BE processor. Results on modern processor architectures and the STI Cell BE are presented.

Program summary

Program title: ITER-REFCatalogue identifier: AECO_v1_0Program summary URL:http://cpc.cs.qub.ac.uk/summaries/AECO_v1_0.htmlProgram obtainable from: CPC Program Library, Queen's University, Belfast, N. IrelandLicensing provisions: Standard CPC licence, http://cpc.cs.qub.ac.uk/licence/licence.htmlNo. of lines in distributed program, including test data, etc.: 7211No. of bytes in distributed program, including test data, etc.: 41 862Distribution format: tar.gzProgramming language: FORTRAN 77Computer: desktop, serverOperating system: Unix/LinuxRAM: 512 MbytesClassification: 4.8External routines: BLAS (optional)Nature of problem: On modern architectures, the performance of 32-bit operations is often at least twice as fast as the performance of 64-bit operations. By using a combination of 32-bit and 64-bit floating point arithmetic, the performance of many dense and sparse linear algebra algorithms can be significantly enhanced while maintaining the 64-bit accuracy of the resulting solution.Solution method: Mixed precision algorithms stem from the observation that, in many cases, a single precision solution of a problem can be refined to the point where double precision accuracy is achieved. A common approach to the solution of linear systems, either dense or sparse, is to perform the LU factorization of the coefficient matrix using Gaussian elimination. First, the coefficient matrix A is factored into the product of a lower triangular matrix L and an upper triangular matrix U. Partial row pivoting is in general used to improve numerical stability resulting in a factorization PA=LU, where P is a permutation matrix. The solution for the system is achieved by first solving Ly=Pb (forward substitution) and then solving Ux=y (backward substitution). Due to round-off errors, the computed solution, x, carries a numerical error magnified by the condition number of the coefficient matrix A. In order to improve the computed solution, an iterative process can be applied, which produces a correction to the computed solution at each iteration, which then yields the method that is commonly known as the iterative refinement algorithm. Provided that the system is not too ill-conditioned, the algorithm produces a solution correct to the working precision.Running time: seconds/minutes     

Torsion classes of <Emphasis Type="Italic">MV</Emphasis>-algebras

 Torsion classes of MV-algebras are defined as radical classes which are closed with respect to homomorphisms; in this paper we investigate their relations to radical classes of lattice ordered groups and to varieties of MV-algebras. Supported by Grant VEGA 1/9056/02.     

Fabrication of buried microfluidic channels with observation windows using femtosecond laser photoablation and parylene-C coating

We developed an advanced method for fabricating microfluidic structures comprising channels and inputs/outputs buried within a silicon wafer based on single level lithography. We etched trenches into a silicon substrate, covered these trenches with parylene-C, and selectively opened their bottoms using femtosecond laser photoablation, forming channels and inputs/outputs by isotropic etching of silicon by xenon difluoride vapors. We subsequently sealed the channels with a second parylene-C layer. Unlike in previously published works, this entire process is conducted at ambient temperature to allow for integration with complementary metal oxide semiconductor devices for smart readout electronics. We also demonstrated a method of chip cryo-cleaving with parylene presence that allows for monitoring of the process development. We also created an observation window for in situ visualization inside the opaque silicon substrate by forming a hole in the parylene layer at the silicon backside and with local silicon removal by xenon difluoride vapor etching. We verified the microfluidic chip performance by forming a segmented flow of a fluorescein solution in an oil stream. This proposed technique provides opportunities for forming simple microfluidic systems with buried channels at ambient temperature.