Keystroke patterns classification using the ARTMAP-FD neural network

This paper presents the development of a keystroke dynamics-based user authentication system using the ARTMAP-FD neural network. The effectiveness of ARTMAPFD in classifying keystroke patterns is analyzed and compared against a number of widely used machine learning systems. The results show that ARTMAP-FD performs well against many of its counterparts in keystroke patterns classification. Apart from that, instead of using the conventional typing timing characteristics, the applicability of typing pressure to ascertaining user's identity is investigated. The experimental results show that combining both latency and pressure patterns can improve the Equal Error Rate (ERR) of the system.

Named entity recognition in unstructured text.

The thesis introduces a set of machine learning techniques that enhance the extraction of Named Entities from informal and unstructured free text.

Learning opportunities for Australian prevocational hospital doctors: exposure, perceived quality and desired methods of learning.

Objective:

To survey prevocational doctors working in Australian hospitals on aspects of postgraduate learning.
Participants and setting:

470 prevocational doctors in 36 health services in Australia, August 2003 to October 2004.
Design:

Cross-sectional cohort survey with a mix of ordinal multicategory questions and free text.
Main outcome measures:

Perceived preparedness for aspects of clinical practice; perceptions of the quantity and usefulness of current teaching and learning methods and desired future exposure to learning methods.
Results:

64% (299/467) of responding doctors felt generally prepared for their job, 91% (425/469) felt prepared for dealing with patients, and 70% (325/467) for dealing with relatives. A minority felt prepared for medicolegal problems (23%, 106/468), clinical emergencies (31%, 146/469), choosing a career (40%, 188/468), or performing procedures (45%, 213/469). Adequate contact with registrars was reported by 90% (418/465) and adequate contact with consultants by 56% (257/466); 20% (94/467) reported exposure to clinical skills training and 11% (38/356) to high-fidelity simulation. Informal registrar contact was described as useful or very useful by 94% (433/463), and high-fidelity simulation by 83% (179/216). Most prevocational doctors would prefer more formal instruction from their registrars (84%, 383/456) and consultants (81%, 362/447); 84% (265/316) want increased exposure to high-fidelity simulation and 81% (283/350) to professional college tutorials.
Conclusion:

Our findings should assist planning and development of training programs for prevocational doctors in Australian hospitals.

Self-rated preparedness of Australian prevocational hospital doctors for emergencies

Objective: To determine perceived preparedness of Australian hospital-based prevocational doctors for resuscitation skills and management of emergencies, and to identify differences between doctors who perceive themselves well prepared and those who perceive themselves poorly prepared for emergencies, in demographics and exposure to desired learning methods.

Methods: Questionnaire consisting of a mix of graded Likert scales and free-text answers distributed to 36 Australian hospitals for secondary distribution to hospital medical officers.

Results: From 2607 questionnaires posted, 470 (18.1%) were returned. Thirty-one per cent (95% confidence interval [CI] 26–35%) felt well prepared for resuscitation and management of emergencies, 41% (CI 37–45%) felt adequately prepared and 28% (CI 24–32%) felt they were not well prepared. Those who felt well prepared reported that they had experienced more exposure to a range of educational methods, including consultant contact, supervisor feedback, clinical skills, high fidelity simulator sessions and unit meetings. Well-prepared and poorly prepared doctors had similar opinions of the usefulness of various learning methods, but the poorly prepared group more frequently expressed a desire for increased exposure to contact with registrars and consultants, clinical skills sessions and hospital and unit meetings. There were no differences in gender, age or country of origin (Australia vs international medical graduates) between those who felt well or poorly prepared.

Conclusions: Many prevocational hospital doctors feel inadequately prepared for the management of emergencies. Perceived preparedness is associated with more exposure to particular educational activities. Increasing exposure to learning of emergencies in undergraduate and prevocational years could reduce the number of junior doctors who feel poorly prepared for emergencies.

You kind of want to fix it don’t you? Exploring general practice trainees’ experiences of managing patients with medically unexplained symptoms

Background
Much of a General Practitioner’s (GP) workload consists of managing patients with medically unexplained symptoms (MUS). GP trainees are often taking responsibility for looking after people with MUS for the first time and so are well placed to reflect on this and the preparation they have had for it; their views have not been documented in detail in the literature. This study aimed to explore GP trainees’ clinical and educational experiences of managing people presenting with MUS.
Method
A mixed methods approach was adopted. All trainees from four London GP vocational training schemes were invited to take part in a questionnaire and in-depth semi-structured interviews. The questionnaire explored educational and clinical experiences and attitudes towards MUS using Likert scales and free text responses. The interviews explored the origins of these views and experiences in more detail and documented ideas about optimising training about MUS. Interviews were analysed using the framework analysis approach.


Results

Eighty questionnaires out of 120 (67 %) were returned and a purposive sample of 15 trainees interviewed. Results suggested most trainees struggled to manage the uncertainty inherent in MUS consultations, feeling they often over-investigated or referred for their own reassurance. They described difficulty in broaching possible psychological aspects and/or providing appropriate explanations to patients for their symptoms. They thought that more preparation was needed throughout their training. Some had more positive experiences and found such consultations rewarding, usually after several consultations and developing a relationship with the patient.
Conclusion
Managing MUS is a common problem for GP trainees and results in a disproportionate amount of anxiety, frustration and uncertainty. Their training needs to better reflect their clinical experience to prepare them for managing such scenarios, which should also improve patient care.

Enhancing email text production for users with motor impairments

People with motor impairments who use a switch device to interface with computers have poor access to affordable software for email communication. The MultiMail email package was developed with government support to provide email access solutions for these users and for others with a range of disabilities. In this paper, the development of accessible on-screen keyboards and a word prediction program which facilitates email text production is discussed. Technology solutions were informed by people with disabilities through focus group and survey data. The resulting cross-disability design of MultiMail provides innovative and cost-free solutions to email text production.

Association of reduced riparian vegetation cover in agricultural landscapes with coarse detritus dynamics in lowland streams

Studies were conducted on streams flowing through agricultural floodplains in south-eastern Australia to quantify whether reductions in riparian canopy cover were associated with alterations to the input and benthic standing stocks of coarse allochthonous detritus. Comparisons were made among three farmland reaches and three reaches within reserves with intact cover of remnant overstorey trees. Detritus inputs to these reaches were measured monthly over 2 years using litter traps. Direct inputs to streams within the reserves were relatively high (550–617 g ash free dry weight (AFDW) m^–2 year^–1), but were lower at farmland reaches with the lowest canopy covers (83–117 gAFDW m^–2 year^–1). Only a minor fraction of the total allochthonous input (<10%) entered any of the study reaches laterally. The mean amounts of benthic detritus were lowest in the most open farmland reaches. Standing stocks of benthic detritus were found to be highly patchy across a large number of agricultural streams, but were consistently very low where the streamside canopy cover was below ~35%. Canopy cover should be restored along cleared agricultural streams because allochthonous detritus is a major source of food and habitat for aquatic ecosystems. Given the absence of pristine lowland streams in south-eastern Australia, those reaches with the most intact remnant overstorey canopies should be used to guide restoration.

Ion and solvent dynamics in gel electrolytes based on ethylene oxide grafted acrylate polymers

Multinuclear pulsed field gradient NMR measurements and rheological viscosity measurements were performed on three series of polymer gel electrolytes. The gels were based on a lithium salt electrolyte swollen into a copolymer matrix comprising an acrylate backbone and ethylene oxide side chains. In each series the side chains differed in length and number, but the acrylate-to-ethylene oxide ratio was kept constant. It was found that the self-diffusion coefficient of the cations was much lower than that of the anions, and that it decreased rapidly when the side chains got longer. In contrast, the self-diffusion coefficient of the anions was found to be independent of chain length. In the gel electrolytes, the diffusion coefficients of the solvent molecules are relatively constant despite an increased viscosity with increasing length of the side chains. However, in salt-free gels made for comparison, the diffusion coefficients of the solvent molecules decreased with increasing length of the side chains, which is consistent with an increased viscosity.

Ion clustering in molecular dynamics simulations of sodium iodide solutions

Model systems of sodium iodide dissolved in dimethyl ether or 1,2-dimethoxyethane (glyme) were studied in order to investigate the structural and dynamic properties of ionic solutions in small and polymeric ethers. Full molecular dynamics simulations were performed at a range of different salt concentrations. An algorithm was designed which assigns ions to clusters and then calculates all the terms which contribute to ionic conductivity. In dilute solutions, free ions are the most common ionic species, followed by ion pairs. As the concentration increases, pairs become the most common species, with significant concentrations of clusters with 3 through 6 ions. Changing the solvent from dimethyl ether to glyme significantly decreases the ion clustering due to the chelate effect in which the two oxygens on a solvent stabilize an associated cation. The conductivity in stable systems is shown to be primarily the result of the movement of free ions and the relative movement of ions within neutral pairs. The Nernst-Einstein relation, commonly used in the discussion of polymer electrolytes, is shown to be inadequate to quantitatively describe conductivity in the model systems.

Free Trade Agreements and US Foreign Policy

The United States has completed numerous free trade agreements (FTAs), but the pattern of these agreements defies conventional explanations. Arguments that are based on domestic interests and economic gains cannot explain the comparative under-performance of US trade agreements. The pattern of US trade agreements is also inconsistent with explanations that focus on state power, which depict FTAs as a “reward” for loyal clients. This article finds a better explanation for the pattern of the United States’ FTAs by consideringthe systemic level of analysis, and in particular the dynamics of the international economic order. It illustrates that strong competition for bilateral trade agreements has resulted in patterns of agreements that the United States cannot easily dominate. This is not to say that the United States has no capacity to finalize trade agreements: the United States remains the world’s most influential nation-state, but the constraints of the international system necessarily limit the degree to which FTAs can serve the interests of US foreign economic policy. The recent evolution of international trade politics, however, indicates that smaller states are comparatively less vulnerable to pressure from great powers, such as the United States.

CISR-ODE, A C plus plus framework with ODE solver for code based system dynamics simulation

Ordinary differential equations are used for modelling a wide range of dynamic systems. Even though there are many graphical software applications for this purpose, a fully customised solution for all problems is code-level programming of the model and solver. In this project, a free and open source C++ framework is designed to facilitate modelling in native code environment and fulfill the common simulation needs of control and many other engineering and science applications. The solvers of this project are obtained from ODEINT and specialised for Armadillo matrix library to provide an easy syntax and a fast execution. The solver code is minimised and its modification for users have become easier. There are several features added to the solvers such as controlling maximum step size, informing the solver about sudden input change and forcing custom times into the results and calling a custom method at these points. The comfort of the model designer, code readability, extendibility and model isolation have been considered in the structure of this framework. The application manages the output results, exporting and plotting them. Modifying the model has become more practical and a portion of corresponding codes are updated automatically. A set of libraries is provided for generation of output figures, matrix hashing, control system functions, profiling, etc. In this paper, an example of using this framework for a classical washout filter model is explained.