K-winner networks.

A special class of mutually inhibitory networks is analyzed, and parameters for reliable K-winner performance are presented. The network dynamics are modeled using interactive activation, and results are compared with the sigmoid model. For equal external inputs, network parameters that select the units with the larger initial activations (the network converges to the nearest stable state) are derived. Conversely, for equal initial activations, networks that select the units with larger external inputs (the network converges to the lowest energy stable state) are derived. When initial activations are mixed with external inputs, anomalous behavior results. These discrepancies are analyzed with several examples. Restrictions on initial states are derived which ensure accurate K-winner performance when unequal external inputs are used.     

An improved method for magnetic identification and localization of cracks in conductors

A SQUID magnetometer can be used to measure the magnetic field produced by flaws in a two-dimensional, conducting plate carrying a current. Identification of the flaw-induced magnetic field is difficult because of the large magnetic field associated with the edges of the plate and the current in the leads that connect the plate to the power supply. We have developed a technique by which the wire and edge fields can be cancelled prior to mapping the magnetic field. In this technique, a similar unflawed conducting sheet is placed adjacent to the flawed plate, with a connection between the sheet and the plate at one edge, and with the opposite edges of the sheet and of the plate connected to the two conductors of a coaxial cable. Thus, an applied current will flow along one conductor of the cable, across the cancelling sheet, cross into the flawed plate, return along the plate, and then return to the power supply along the other conductor of the coaxial cable. As a result of this geometry, there is no magnetic field from the lead-in wires because they are coaxial, and the magnetic field due to the edges of the plate is cancelled by the opposing magnetic field of the edges in the adjacent sheet. The extent of cancellation is determined primarily by the separation between the plate and the cancelling sheet, by the thickness of the plate, and by macroscopic inhomogeneities in their electrical conductivities.     

Dependence of gate oxide breakdown on initial charge trapping under Fowler-Nordheim injection

This work demonstrates that for constant oxide reliability stresses in the Fowler-Nordheim regime a low initial rate of charge trapping/detrapping results in long times to breakdown. It was found for MOS gate oxides that when the initial trapping has been completed at low fields times to breakdown enhance. Depending on the stress sequence measurement results can vary significantly which is of great relevance for correct oxide lifetime predictions.     

Experimental evaluation of a multivariable adaptive nonlinear predictive controller

A new multivariable adaptive nonlinear predictive controller is designed using a general nonlinear input-output model and variable transformations. The controller is similar in form to typical linear predictive controllers can be tuned analogously or by specifying a single parameters for each controlled variable. In addition, the design procedure is computationally efficient. The new controller is compared to a multi-loop proportional-integral (PI) controller with one-way static decoupling and to an adaptive linear predictive controller through tests on a simulated nonlinear distillation column. The new controller performed well in an experimental application to a multicomponent distillation column.     

Technique for Producing Mullite and Other Mixed-Oxide Systems

Mixed-oxide systems are of interest in a variety of technical ceramic applications. The key in many systems is to achieve a high degree of homogeneity, approaching the molecular level. A technique for producing mixed oxides via direct oxidation of mixtures of organometallic compounds is discussed as a potential route. Preliminary work performed on producing mullite is discussed.     

Hemispheric priming by meaningful and nonmeaningful symbols in language-trained chimpanzees (Pan troglodytes): Further evidence of a left hemisphere advantage.

Hemispheric priming was examined in 3 language-trained chimpanzees (Pan troglodytes) using a simple reaction time (RT) paradigm. Ss were required to hold down a response button until the occurrence of a response cue. A warning stimulus was presented to either the left visual field or the right visual field (RVF) before the response cue occurred. No warning stimulus was presented on control trials. The warning stimuli were geometric communicative symbols from 2 semantic categories: food and tools. A 3rd set of warning stimuli were familiar geometric symbols. Dependent measures included RT and the number of false-positive responses. RT data indicated an RVF advantage in priming when the warning stimuli were food or tool symbols. No significant visual half-field differences were found for familiar symbols, but a trend toward an RVF advantage was observed. These effects were enhanced when Ss responded with their left hand. False-positive data also indicated an RVF advantage for the food and tool warning stimuli. The data indicate that hemispheric asymmetries for processing communicative symbols are present in language-trained chimpanzees. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Cognitive categorization and quality of performance ratings.

The effects of cognitive categorization of raters on accuracy, leniency, and halo of performance evaluations were investigated in a field setting. One hundered seventy-four subordinates evaluated the performance of their managers on three performance dimensions. Managers were categorized as congruent or incongruent based on subordinates' perceptions of the extent to which the manager's behavior met the subordinates' expectations. The results indicated that the quality of ratings assigned by subordinates was related to the cognitive categories used. As hypothesized, ratings of managers who were categorized as congruent were found to be more accurate and also to contain more leniency and halo tendency than the ratings of managers who were categorized as incongruent. Implications of these findings for performance-appraisal research are discussed. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

An Overview of Next-Generation Androgen Receptor-Targeted Therapeutics in Development for the Treatment of Prostate Cancer

Traditional endocrine therapy for prostate cancer (PCa) has been directed at suppression of the androgen receptor (AR) signaling axis since Huggins et al. discovered that diethylstilbestrol (DES; an estrogen) produced chemical castration and PCa tumor regression. Androgen deprivation therapy (ADT) still remains the first-line PCa therapy. Insufficiency of ADT over time leads to castration-resistant PCa (CRPC) in which the AR axis is still active, despite castrate levels of circulating androgens. Despite the approval and use of multiple generations of competitive AR antagonists (antiandrogens), antiandrogen resistance emerges rapidly in CRPC due to several mechanisms, mostly converging in the AR axis. Recent evidence from multiple groups have defined noncompetitive or noncanonical direct binding sites on AR that can be targeted to inhibit the AR axis. This review discusses new developments in the PCa treatment paradigm that includes the next-generation molecules to noncanonical sites, proteolysis targeting chimera (PROTAC), or noncanonical N-terminal domain (NTD)-binding of selective AR degraders (SARDs). A few lead compounds targeting each of these novel noncanonical sites or with SARD activity are discussed. Many of these ligands are still in preclinical development, and a few early clinical leads have emerged, but successful late-stage clinical data are still lacking. The breadth and diversity of targets provide hope that optimized noncanonical inhibitors and/or SARDs will be able to overcome antiandrogen-resistant CRPC.     

Photodegradation of ethylene–propylene copolymer and ethylene–propylene–ethylidenenorbornene terpolymer

Significant progress has been made in recent years regarding the photooxidation of olefin copolymers, but questions still remain. This paper reviews the progress and probes the photooxidative chemistry of ethylene–propylene (EP) and ethylene/propylene/diene monomer (EPDM) copolymers. Both stabilized and unstabilized polymer plaques were irradiated in a xenon are and the surface chemistry followed using infra-red spectroscopy. Model compounds were used to help elucidate the chemistry caused by unique structural features present in the copolymers. Volatile products evolved during photooxidation were determined giving valuable insight into the degradation chemistry.     

Inductive transfer with context-sensitive neural networks

Context-sensitive Multiple Task Learning, or csMTL, is presented as a method of inductive transfer which uses a single output neural network and additional contextual inputs for learning multiple tasks. Motivated by problems with the application of MTL networks to machine lifelong learning systems, csMTL encoding of multiple task examples was developed and found to improve predictive performance. As evidence, the csMTL method is tested on seven task domains and shown to produce hypotheses for primary tasks that are often better than standard MTL hypotheses when learning in the presence of related and unrelated tasks. We argue that the reason for this performance improvement is a reduction in the number of effective free parameters in the csMTL network brought about by the shared output node and weight update constraints due to the context inputs. An examination of IDT and SVM models developed from csMTL encoded data provides initial evidence that this improvement is not shared across all machine learning models.