Effects of curtailment on an admissions model for a graduate management program.

Developed 2 admissions models to predict future academic performance of graduate management students. The 1st model was based on 40 students who were admitted and enrolled in the program and was uncorrected for curtailment. The 2nd model was developed from the total applicant population of 222 students after curtailment correction. The corrected model demonstrated higher predictive validity than the uncorrected model for 2 future classes of students. Furthermore, different predictors entered each model, affecting the beta weights, validities of predictors, and the total amount of variance explained by the models. A factor analysis and an analysis of admissions decisions offer additional support for the curtailment-corrected model in selecting students with high academic potential. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Visualization of the arteries, veins, and nerves of the orbit by sector computed tomography

Altering the scintillation crystal configuration of a fixed-detector CT scanner increased the spatial resolution from 6.25 to 10 line pairs per centimeter. Sector scans of th orbit with this modified scanner showed the ophthalmic artery and superior ophthalmic vein and their branches and the intraorbital branches of the third, fifth, and sixth nerves, apparently seen with CT for the first time.     

Does increased confidence among novice drivers imply a decrease in safety? The effects of skid training on slippery road accidents

Finnish driver training was renewed in 1990 with the inclusion of a compulsory skid training course in the curriculum. The study evaluated the renewal's effect on accidents in slippery road conditions. A questionnaire was sent by mail to 41000 novice drivers who were randomly selected from the official register of driving licences. It included questions on driving exposure and the accidents the drivers had been involved in during 6-18 months following licensing. The rate of return was 74.7%. Half of the drivers had received their licence in 1989 and had, therefore, not received any skid training. The other half had received their licence in 1990 after the introduction of the skid training course. The results showed no effects of the renewal on slippery road accidents for either male or female drivers. Another questionnaire was sent to 1300 old and new curriculum drivers immediately after licensing and a second time 1/2-1 year later, both with questions about skills, worries and perceived risks regarding driving in slippery conditions. The new curriculum drivers showed higher confidence in their skills and they were less afraid to drive in slippery conditions than the old curriculum drivers. This increase in confidence as a result of skid training is discussed. It is argued that high confidence in one's personal skills does not necessarily imply negative safety. The crucial factor is how these skills are used, and for what purpose.     

Approximating the Advertisement Placement Problem

The advertisement placement problem deals with space and time sharing by advertisements on the Internet. Consider a Web page containing a rectangular display area (e.g., a banner) in which advertisements may appear. The display area can be utilized efficiently by allowing several small ads to appear simultaneously side by side, as well as by cycling through a schedule of ads, allowing different ads to be displayed at different times. A customer wishing to purchase advertising space specifies an ad size and a display count, which is the number of times their ad should appear during each cycle. The scheduler may accept or reject any given advertisement, but must be able to schedule all accepted ads within the given time and space constraints. Each advertisement has a non-negative profit associated with it, and the objective is to schedule a maximum-profit subset of ads. We present a (3 + )-approximation algorithm for the general problem, as well as (2 + )-approximation algorithms for two special cases.     

A Simple Technique for Enhancing the Hypothesis Testing Capability of Devices Sensitive to Phase in the Presence of Narrow-Band Normal Noise

A new statistic is described for hypothesis testing on phase in the presence of narrow-band normal noise when the true phase, itself, is a nonuniformly distributed random variable. Comparisons are made with the more standard direct techniques.     

Financing cleaner production investments—UNEP experience

Cleaner production (CP) can be good business. To address constraints related to financing CP and to look for ways to overcome them, the United Nations Environment Programme, Division of Technology, Industry and Economic (UNEP/DTIE), has been implementing since 1999 a project "Strategies and mechanisms for promoting cleaner production investment in developing countries". The project is coming to an end in February 2003. Through field testing and adaptation and in partnership with experts in industrialised and developing countries and economies in transition, it has generated a series of guides, checklists and training packages. Over 3,000 persons attended the over 100 1- or 2-day courses delivered by national institutions in five demonstration countries. The project has also contributed to making preventive approaches and efficient resource management part of awareness raising and training activities by some development finance institutions and the UNEP Finance Initiatives. Proposals on innovative financing schemes have been made in several workshops for banks and donors in 2002. In addition to the special financing issue of the Journal of Cleaner Production (published in March 2003), articles on links between CP, financing and accounting have been contributed to professional publications.Although the project's demonstration activities have focused on five countries in Africa, Asia and Central America, its conclusions, recommendations and outputs are of direct relevance also to other developing countries and economies in transition. The integration of preventive strategies into investment decision making processes is a challenge both in the industrialised and in developing countries in the region. This paper summarises the lessons learned through the project and presents the recommendations, tools and instruments available for adaptation and use by CP promoters, financial institutions and business schools.     

On the determination of the post-irradiation time from the glow curve of TLD-100

The ratio of peak 3 to the sum of peaks 4 + 5 in TLD-100 was measured for various pre-irradiation and post-irradiation time periods, under conditions characteristic of routine personal dosimetry. It was confirmed that the value of this ratio depends only on the elapsed time between the prior readout and the present one, independent of the moment when the irradiation took place during the total time interval (storage time). This effect indicates that fading of peak 3 seems to be due mainly to changes in the unoccupied traps, and not to decay of trapped charges, being almost independent of the presence of electrons or holes in the traps. This observation leads to the conclusion that the suggestions in the past to use the decay of peak 3 in TLD-100 for the measurement of the elapsed time between irradiation and readout may have been wrong. On the other hand, the decay of peak 2 can be used to measure the elapsed time from irradiation, since the rate of decay is different when related to pre-irradiation and post-irradiation times, indicating a much higher decay rate of the trapped charges (Randall-Wilkins decay). However, because of the fast decay rate of peak 2, its use for determination of the elapsed time since irradiation is of little practical significance.     

Pricing based adaptive scheduling method for bandwidth allocation

In this paper we present a packet scheduling method which guarantees bandwidth of the connection and optimizes revenue of the network service provider. A closed form formula for updating the adaptive weights of a packet scheduler is derived from a revenue-based optimization problem. The weight updating procedure is fast and independent on the assumption of the connections’ statistical behavior. The features of the algorithm are simulated and analyzed with and without a call admission control (CAC) mechanism. We also show in context with the CAC procedure a mechanism for guaranteeing a specified mean bandwidth for different service classes.     

Interlingua-based English–Korean Two-way Speech Translation of Doctor–Patient Dialogues with CCLINC

Development of a robust two-way real-time speech translationsystem exposes researchers and system developers to various challenges of machine translation(MT) and spoken language dialogues. The need for communicating in at least two differentlanguages poses problems not present for a monolingual spoken language dialogue system,where no MT engine is embedded within the process flow. Integration of various componentmodules for real-time operation poses challenges not present for text translation. In this paper,we present the CCLINC (Common Coalition Language System at Lincoln Laboratory) English–Koreantwo-way speech translation system prototype trained on doctor–patient dialogues,which integrates various techniques to tackle the challenges of automatic real-time speechtranslation. Key features of the system include (i) language–independent meaning representation which preserves the hierarchicalpredicate–argument structure of an input utterance, providing a powerful mechanism for discourse understanding of utterances originating from different languages,word-sense disambiguation and generation of various word orders of many languages, (ii) adoptionof the DARPA Communicator architecture, a plug-and-play distributed system architecturewhich facilitates integration of component modules and system operation in real time, and (iii)automatic acquisition of grammar rules and lexicons for easy porting of the system to differentlanguages and domains. We describe these features in detail and present experimental results.     

Prediction of Residual Stresses Using Partial Least Squares Regression on Barkhausen Noise Signals

Residual stresses are quantitatively predicted based on the Barkhausen noise measurement with a partial least squares regression model. The measurements are taken from two sets of case-hardened samples. The benefits of using certain feature elimination strategies prior model identification are also studied. The elimination methods applied are correlation-based elimination, uninformative variable elimination and successive projections algorithm. The results show that the best predictions are usually obtained when the successive projections algorithm is applied. The prediction accuracy of the best models found shows that partial least squares models can be successfully used for prediction of material properties based on the Barkhausen noise measurement.