A tutorial on the cross-entropy method

The cross-entropy (CE) method is a new generic approach to combinatorial and multi-extremal optimization and rare event simulation. The purpose of this tutorial is to give a gentle introduction to the CE method. We present the CE methodology, the basic algorithm and its modifications, and discuss applications in combinatorial optimization and machine learning. combinatorial optimization

Aboriginal and Torres Strait Islander Studies Unit Annual Report

Welcome to the 2002 Aboriginal and Torres Strait Islander Studies Unit Annual Report. This report is a brief summary of Unit activities during the 2002 calendar year. The Unit provides personal and academic support for Aboriginal and Torres Strait Islander students and specifically aims to increase the recruitment, retention, academic performance and graduation rates of Indigenous students. The Unit also administers schemes to help Indigenous students gain access to, and receive support in, tertiary studies such as the Alternative Entry scheme and the federally funded Aboriginal Tutorial Assistance Scheme (ATAS). The Unit is also the focus for teaching and research in Aboriginal and Torres Strait Islander Studies at the University of Queensland.

Aboriginal and Torres Strait Islander Studies Unit Annual Report 2003

Welcome to the 2003 Aboriginal and Torres Strait Islander Studies Unit Annual Report. This report is a brief summary of Unit activities during the 2003 calendar year. The Unit provides personal and academic support for Aboriginal and Torres Strait Islander students and specifically aims to increase the recruitment, retention, academic performance and graduation rates of Indigenous students. The Unit also administers schemes to help Indigenous students gain access to, and receive support in, tertiary studies such as the Alternative Entry scheme and the federally funded Aboriginal Tutorial Assistance Scheme (ATAS). The Unit is also the focus for teaching and research in Aboriginal and Torres Strait Islander Studies at the University of Queensland.

Aboriginal and Torres Strait Islander Studies Unit Annual Report 2005

Welcome to the 2005 Aboriginal and Torres Strait Islander Studies Unit Annual Report. This report is a brief summary of Unit activities during the 2005 calendar year. The Unit provides personal and academic support for Aboriginal and Torres Strait Islander students and specifically aims to increase the recruitment, retention, academic performance and graduation rates of Indigenous students. The Unit also administers schemes to help Indigenous students gain access to, and receive support in, tertiary studies such as the Alternative Entry scheme and the federally-funded Indigenous Tutorial Assistance Scheme (ITAS). The Unit is also the focus for teaching and research in Aboriginal and Torres Strait Islander Studies at the University of Queensland.

Hypothalamic dopamine D1 receptors are involved in the stimulation of prolactin secretion by high environmental temperature in the female sheep

Recent evidence suggests that dopamine, acting via its D1 receptors, may function as a neurotransmitter in intrahypothalamic pathways involved in the stimulation of prolactin secretion. Functional dopamine D1 receptors are present in the ventromedial hypothalamic nucleus (VMH) and we hypothesized that they might be part of a prolactin-stimulatory pathway activated by stress. We tested this hypothesis in a series of experiments on sheep involving two different forms of stressors, audiovisual (barking dog) and high environmental temperature. We attempted to block the stimulation of prolactin secretion by infusion into the VMH of an antagonist specific for the D1 receptor. Ovariectomised, oestradiol-implanted merino ewes were surgically implanted with bilateral guide tubes directed at the VMH. After a 180 min pretreatment period, the ewes either were or were not exposed to a stressor (30 min of barking dog or 120 min at 35 degrees C, 65% relative humidity). D1 receptor antagonist, SCH23390 or vehicle (0.9% saline) was infused into the VMH (1.7 mu l/h, 120 nmol/h) for 60 min prior to and during the stressor period. Blood was sampled every 15 min via jugular cannulae and the plasma was assayed for prolactin, cortisol and growth hormone (GH). Both stressors significantly increased prolactin concentrations over control levels. SCH23390 infusion significantly attenuated the prolactin response to high environmental temperature, but had no effect on the prolactin response to audiovisual stress. Cortisol concentrations were significantly increased by audiovisual stress only and were not affected by SCH23390, GH concentrations were not changed by either stressor or infusion. Drug infusion alone did not affect the concentration of the hormones. The data suggest that the VMH D1 receptors are involved in a prolactin stimulatory pathway in response to high environmental temperature. The inability of the D1 antagonist to affect the response to the barking dog indicates that this pathway is stress-specific, implying that there is more than one mechanism or pathway involved in the prolactin response to different stressors.

Developing logic programs from specifications using stepwise refinement

In this paper we demonstrate a refinement calculus for logic programs, which is a framework for developing logic programs from specifications. The paper is written in a tutorial-style, using a running example to illustrate how the refinement calculus is used to develop logic programs. The paper also presents an overview of some of the advanced features of the calculus, including the introduction of higher-order procedures and the refinement of abstract data types.

Use of the web to provide learning support for a large metabolism and nutrition class

To support student learning in a large Metabolism and Nutrition class, we have introduced a web-based package, using a commercially available program, WebCT. The package was developed at a minimal cost and with limited resources. In addition to downloadable (PDF) versions of lecture Powerpoint presentations, tutorial outlines and a practical class exercise, web-based self-directed learning exercises were included to reinforce and extend lecture material in an active learning environment. The web-site also contained a variety of formative and summative assessment tasks that examined both factual recall and higher order thinking Detailed course information, timetables and a bulletin board were also readily accessible. Student usage of the site was generally high, but varied widely between individual students. Students who achieved a high overall score for the course completed on average three times as many formative assessment items and achieved a higher score for all tests than students who did poorly. Student feedback about the site was very positive with the majority of students reporting that the course material and assessment items that were available were useful to their learning. Administration of the course was also facilitated. (C) 2001 IUBMB. Published by Elsevier Science Ltd. All rights reserved.

The insulin hypoglycemia test: Hypoglycemic criteria and reproducibility

The insulin hypoglycemia test (IHT) is widely regarded as the 'gold standard' for dynamic stimulation of the hypothalamic-pituitary-adrenal (HPA) axis. This study aimed to investigate the temporal relationship between a rapid decrease in plasma glucose and the corresponding rise in plasma adenocorticotropic hormone (ACTH), and to assess the reproducibility of hormone responses to hypoglycemia in normal humans. Ten normal subjects underwent IHTs, using an insulin dose of 0.15 U/kg. Of these, eight had a second IHT (IHT2) and three went on to a third test (IHT3). Plasma ACTH and cortisol were measured at 15-min intervals and, additionally, in four IHT2s and the three IHT3s, ACTH was measured at 2.5- or 5-min intervals. Mean glucose nadirs and mean ACTH and cortisol responses were not significantly different between IHT1, IHT2 and IHT3. Combined data from all 21 tests showed the magnitude of the cortisol responses, but not the ACTH responses, correlated significantly with the depth and duration of hypoglycemia. All subjects achieved glucose concentrations of of less than or equal to 1.6 mmol/l before any detectable rise in ACTH occurred. In the seven tests performed with frequent sampling, an ACTH rise never preceeded the glucose nadir, but occurred at the nadir, or up to 15 min after. On repeat testing, peak ACTH levels varied markedly within individuals, whereas peak cortisol levels were more reproducible (mean coefficient of variation 7%). In conclusion, hypoglycemia of less than or equal to 1.6 mmol/l was sufficient to cause stimulation of the HPA axis in all 21 IHTs conducted in normal subjects. Nonetheless; our data cannot reveal whether higher glucose nadirs would stimulate increased HPA axis activity in all subjects. Overall, the cortisol response to hypoglycemia is more reproducible than the ACTH response but, in an individual subject, the difference in peak cortisol between two IHTs may exceed 100 nmol/l.

Rumors and stable-cause attribution in prediction and behavior

Two stock-market simulation experiments investigated the notion that rumors that invoke stable-cause attributions spawn illusory associations and less regressive predictions and behavior. In Study 1, illusory perceptions of association and stable causation (rumors caused price changes on the day after they appeared) existed despite rigorous conditions of nonassociation (price changes were unrelated to rumors). Predictions (recent price trends will continue) and trading behavior (departures from a strong buy-low-sell-high strategy) were both anti-regressive. In Study 2, stability of attribution was manipulated via a computerized tutorial. Participants taught to view price-changes as caused by stable forces predicted less regressively and departed more from buy-low-sell-high trading patterns than those taught to perceive changes as caused by unstable forces. Results inform a social cognitive and decision theoretic understanding of rumor by integrating it with causal attribution, covariation detection, and prediction theory. (C) 2002 Elsevier Science (USA). All rights reserved.