Precision-recall operating characteristic (P-ROC) curves in imprecise environments

Traditionally, machine learning algorithms have been evaluated in applications where assumptions can be reliably made about class priors and/or misclassification costs. In this paper, we consider the case of imprecise environments, where little may be known about these factors and they may well vary significantly when the system is applied. Specifically, the use of precision-recall analysis is investigated and compared to the more well known performance measures such as error-rate and the receiver operating characteristic (ROC). We argue that while ROC analysis is invariant to variations in class priors, this invariance in fact hides an important factor of the evaluation in imprecise environments. Therefore, we develop a generalised precision-recall analysis methodology in which variation due to prior class probabilities is incorporated into a multi-way analysis of variance (ANOVA). The increased sensitivity and reliability of this approach is demonstrated in a remote sensing application.

Dynamics of a mesoscopic charge quantum bit under continuous quantum measurement

We present the conditional quantum dynamics of an electron tunneling between two quantum dots subject to a measurement using a low transparency point contact or tunnel junction. The double dot system forms a single qubit and the measurement corresponds to a continuous in time readout of the occupancy of the quantum dot. We illustrate the difference between conditional and unconditional dynamics of the qubit. The conditional dynamics is discussed in two regimes depending on the rate of tunneling through the point contact: quantum jumps, in which individual electron tunneling current events can be distinguished, and a diffusive dynamics in which individual events are ignored, and the time-averaged current is considered as a continuous diffusive variable. We include the effect of inefficient measurement and the influence of the relative phase between the two tunneling amplitudes of the double dot/point contact system.

Genetic assignment methods for the direct, real-time estimation of migration rate: a simulation-based exploration of accuracy and power

Genetic assignment methods use genotype likelihoods to draw inference about where individuals were or were not born, potentially allowing direct, real-time estimates of dispersal. We used simulated data sets to test the power and accuracy of Monte Carlo resampling methods in generating statistical thresholds for identifying F-0 immigrants in populations with ongoing gene flow, and hence for providing direct, real-time estimates of migration rates. The identification of accurate critical values required that resampling methods preserved the linkage disequilibrium deriving from recent generations of immigrants and reflected the sampling variance present in the data set being analysed. A novel Monte Carlo resampling method taking into account these aspects was proposed and its efficiency was evaluated. Power and error were relatively insensitive to the frequency assumed for missing alleles. Power to identify F-0 immigrants was improved by using large sample size (up to about 50 individuals) and by sampling all populations from which migrants may have originated. A combination of plotting genotype likelihoods and calculating mean genotype likelihood ratios (D-LR) appeared to be an effective way to predict whether F-0 immigrants could be identified for a particular pair of populations using a given set of markers.

Why do eastern curlews Numenius madagascariensis feed on prey that lowers intake rate before migration?

Eastern curlews Numenius madagascariensis spending the nonbreeding season in eastern Australia foraged on three intertidal decapods: soldier crab Mictyris longicarpus, sentinel crab Macrophthalmus crassipes and ghost-shrimp Trypaea australiensis. Due to their ecology, these crustaceans were spatially segregated (=distributed in 'patches') and the curlews intermittently consumed more than one prey type. It was predicted that if the curlews behaved as intake rate maximizers, the time spent foraging on a particular prey (patch) would reflect relative availabilities of the prey types and thus prey-specific intake rates would be equal. During the mid-nonbreeding period (November-December), Mictyris and Macrophthalmus were primarily consumed and prey-specific intake rates were statistically indistinguishable (8.8 versus 10.1 kJ x min(-1)). Prior to migration (February), Mictyris and Trypaea were hunted and the respective intake rates were significantly different (8.9 versus 2.3 kJ x min(-1)). Time allocation to Trypaea-hunting was independent of the availability of Mictyris. Thus, consumption of Trypaea depressed the overall intake rate. Six hypotheses for consuming Trypaea before migration were examined. Five hypotheses: the possible error by the predator, prey specialization, observer overestimation of time spent hunting Trypaea, supplementary prey and the choice of higher quality prey due to a digestive bottleneck, were deemed unsatisfactory. The explanation for consumption of a low intake-rate but high quality prey (Trypaea) deemed plausible was diet optimisation by the Curlews in response to the pre-migratory modulation (decrease in size/processing capacity) of their digestive system. With a seasonal decrease in the average intake rate, the estimated intake per low tide increased from 1233 to 1508 kJ between the mid-nonbreeding and pre-migratory periods by increasing the overall time spent on the sandflats and the proportion of time spent foraging.

Improving precision and reducing bias in biological surveys: estimating false-negative error rates

The use of presence/absence data in wildlife management and biological surveys is widespread. There is a growing interest in quantifying the sources of error associated with these data. We show that false-negative errors (failure to record a species when in fact it is present) can have a significant impact on statistical estimation of habitat models using simulated data. Then we introduce an extension of logistic modeling, the zero-inflated binomial (ZIB) model that permits the estimation of the rate of false-negative errors and the correction of estimates of the probability of occurrence for false-negative errors by using repeated. visits to the same site. Our simulations show that even relatively low rates of false negatives bias statistical estimates of habitat effects. The method with three repeated visits eliminates the bias, but estimates are relatively imprecise. Six repeated visits improve precision of estimates to levels comparable to that achieved with conventional statistics in the absence of false-negative errors In general, when error rates are less than or equal to50% greater efficiency is gained by adding more sites, whereas when error rates are >50% it is better to increase the number of repeated visits. We highlight the flexibility of the method with three case studies, clearly demonstrating the effect of false-negative errors for a range of commonly used survey methods.

High-fidelity Z-measurement error encoding of optical qubits

We demonstrate a quantum error correction scheme that protects against accidental measurement, using a parity encoding where the logical state of a single qubit is encoded into two physical qubits using a nondeterministic photonic controlled-NOT gate. For the single qubit input states vertical bar 0 >, vertical bar 1 >, vertical bar 0 > +/- vertical bar 1 >, and vertical bar 0 > +/- i vertical bar 1 > our encoder produces the appropriate two-qubit encoded state with an average fidelity of 0.88 +/- 0.03 and the single qubit decoded states have an average fidelity of 0.93 +/- 0.05 with the original state. We are able to decode the two-qubit state (up to a bit flip) by performing a measurement on one of the qubits in the logical basis; we find that the 64 one-qubit decoded states arising from 16 real and imaginary single-qubit superposition inputs have an average fidelity of 0.96 +/- 0.03.

Lingual kinematic strategies used to increase speech rate: Comparisons between younger and older adults

The primary objective of this study was to assess the lingual kinematic strategies used by younger and older adults to increase rate of speech. It was hypothesised that the strategies used by the older adults would differ from the young adults either as a direct result of, or in response to a need to compensate for, age-related changes in the tongue. Electromagnetic articulography was used to examine the tongue movements of eight young (M526.7 years) and eight older (M567.1 years) females during repetitions of /ta/ and /ka/ at a controlled moderate rate and then as fast as possible. The younger and older adults were found to significantly reduce consonant durations and increase syllable repetition rate by similar proportions. To achieve these reduced durations both groups appeared to use the same strategy, that of reducing the distances travelled by the tongue. Further comparisons at each rate, however, suggested a speed-accuracy trade-off and increased speech monitoring in the older adults. The results may assist in differentiating articulatory changes associated with normal aging from pathological changes found in disorders that affect the older population.

Blood pressure and heart rate during tonic immobility in the black tipped reef shark, Carcharhinus melanoptera

Tonic immobility was induced in black tipped reef sharks (Carcharhinus melanoptera) and heart rate and ventral aortic blood pressure recorded. Without branchial irrigation, tonic immobility was correlated with a significant depression in blood pressure and heart rate irrespective of the sharks being in air or in water. Tonic immobility with branchial irrigation resulted in a significant increase in blood pressure in sharks in air, but not in water. Heart rate was unchanged when the gills were irrigated. Intra-arterial injections of atropine abolished the bradycardia and blood pressure rise associated with tonic immobility. We conclude that, during tonic immobility, sharks are able to receive afferent information from the ventilatory system and make appropriate responses via the vagus nerve.

Relationships between blood pressure and heart rate in the saltwater crocodile Crocodylus porosus

The cardiac limb of the baroreflex loop was studied in the saltwater crocodile Crocodylus porosus, The classical pharmacological methodology using phenylephrine and sodium nitroprusside was used to trigger blood pressure changes, and the resulting alterations in heart rate were analysed quantitatively using a logistic function. Interindividual differences in resting heart rates and blood pressures were observed, but all seven animals displayed clear baroreflex responses. Atropine and sotalol greatly attenuated the response. A maximal baroreflex gain of 7.2 beats min(-1) kPa(-1) was found at a mean aortic pressure of 6.1 kPa, indicating the active role of the baroreflex in a wide pressure range encompassing hypotensive and hypertensive states. At the lowest mean aortic pressures (5.0 kPa), the synergistic role of the pulmonary-to-systemic shunt in buffering the blood pressure drop also contributes to blood pressure regulation, Pulse pressure showed a better correlation,vith heart rate and also a higher gain than mean aortic, systolic or diastolic pressures, and this is taken as an indicator of the existence of a differential control element working simultaneously with a linear proportional element.

Heart rate, blood lactate and kinematic data of elite colts (under-19) rugby union players during competition

Physiological and kinematic data were collected from elite under-19 rugby union players to provide a greater understanding of the physical demands of rugby union. Heart rate, blood lactate and time-motion analysis data were collected from 24 players (mean +/- s((x) over bar): body mass 88.7 +/- 9.9 kg, height 185 +/- 7 cm, age 18.4 +/- 0.5 years) during six competitive premiership fixtures. Six players were chosen at random from each of four groups: props and locks, back row forwards, inside backs, outside backs. Heart rate records were classified based on percent time spent in four zones (>95%, 85-95%, 75-84%, <75% HRmax). Blood lactate concentration was measured periodically throughout each match, with movements being classified as standing, walking, jogging, cruising, sprinting, utility, rucking/mauling and scrummaging. The heart rate data indicated that props and locks (58.4%) and back row forwards (56.2%) spent significantly more time in high exertion (85-95% HRmax) than inside backs (40.5%) and outside backs (33.9%) (P < 0.001). Inside backs (36.5%) and outside backs (38.5%) spent significantly more time in moderate exertion (75-84% HRmax) than props and locks (22.6%) and back row forwards (19.8%) (P < 0.05). Outside backs (20.1%) spent significantly more time in low exertion (< 75% HRmax) than props and locks (5.8%) and back row forwards (5.6%) (P < 0.05). Mean blood lactate concentration did not differ significantly between groups (range: 4.67 mmol.l(-1) for outside backs to 7.22 mmol.l(-1) for back row forwards; P < 0.05). The motion analysis data indicated that outside backs (5750 m) covered a significantly greater total distance than either props and locks or back row forwards (4400 and 4080 m, respectively; P < 0.05). Inside backs and outside backs covered significantly greater distances walking (1740 and 1780 m, respectively; P < 0.001), in utility movements (417 and 475 m, respectively; P < 0.001) and sprinting (208 and 340 m, respectively; P < 0.001) than either props and locks or back row forwards (walking: 1000 and 991 m; utility movements: 106 and 154 m; sprinting: 72 and 94 m, respectively). Outside backs covered a significantly greater distance sprinting than inside backs (208 and 340 m, respectively; P < 0.001). Forwards maintained a higher level of exertion than backs, due to more constant motion and a large involvement in static high-intensity activities. A mean blood lactate concentration of 4.8-7.2 mmol.l(-1) indicated a need for 'lactate tolerance' training to improve hydrogen ion buffering and facilitate removal following high-intensity efforts. Furthermore, the large distances (4.2-5.6 km) covered during, and intermittent nature of, match-play indicated a need for sound aerobic conditioning in all groups (particularly backs) to minimize fatigue and facilitate recovery between high-intensity efforts.

Increased ovulation rate in Merino ewes immunized against small synthetic peptide fragments of the inhibin alpha subunit

Four experiments were carried out in Merino ewes during a period of 4 years to determine the long-term effects of immunization against different synthetic peptides mimicking the amine terminal of the or subunit of porcine inhibin. Peptides were conjugated to human serum albumin and 100-200 mu g emulsified in Freund's complete adjuvant for the primary immunization. Usually two booster injections were given at monthly intervals with 50-100 mu g conjugated peptide using either incomplete Freund's adjuvant or Montanide : Marcel. In some experiments a further immunization was carried in the next year. Blood samples were taken 10 days after each immunization, during the luteal phase, for estimation of gonadotrophin concentrations and determination of inhibin antibody titres. One day after blood sampling cloprostenol was used to induce luteolysis and laparoscopy was performed in the subsequent oestrous cycle. Immunization of ewes with synthetic peptides 1-32, 1-26, 7-26 and 8-30 resulted in large increases in the ovulation rate (OR). An approximately two-fold increase in OR was observed following the first booster immunization with these peptides and a three- to five-fold increase after the second booster immunization. Immunization with these large peptides resulted in a sustained increase in OR for a period of at least 1 year after the second booster immunization. Of the shorter peptides, peptides 10-26 and 13-26 gave a reasonable ovulatory response, although it was more difficult to obtain a response with peptides 1-16, 8-22, 13-25, 8-19 and 10-19; peptides 7-13 and 1-6 gave no response (but were examined for one breeding season only). The smaller peptides led to lower inhibin antibody titres that were not necessarily associated with increased follicle-stimulating hormone (FSH) or OR. More intensive blood sampling in one experiment showed that following primary immunization against peptide 1-32 there was a transient increase in plasma FSH which did not lead to an increased OR. Moreover, a prolonged period of raised FSH after the first booster was significantly correlated with increased OR. In these animals antibody titres were only slightly increased after primary immunization, but after the first booster immunization higher titres were observed that were significantly correlated with trough FSH values and the subsequent OR. These results are interpreted as showing that (1) to obtain an increase in OR peptides 1-32, 1-26 and 7-26 are suitable as immunogens; (2) smaller peptides are less reliable, often require multiple injections, and the response may be delayed; and (3) an extended period of raised plasma FSH is needed to give a large ovulatory response.

Between-year fluctuations in schizophrenia birth rate and associations with the Southern Oscillation Index, cloud cover and sunshine

Our group have recently proposed that low prenatal vitamin D may be a risk-modifying factor for schizophrenia. Climate variability impacts on vitamin D levels in a population via fluctuations in the amount of available UV radiation. In order to explore this hypothesis, we examined fluctuations in the birthrates for people with schizophrenia born between 1920 and 1967 and three sets of variables strongly associated with UV radiation. These included: (a) the Southern Oscillation Index (SOI), a marker of El Nino which is the most prominent meteorological factor that influences Queensland weather: (b) measures of cloud cover and (c) measures of sunshine. Schizophrenia births were extracted from the Queensland Mental Health register and corrected for background population birth rates. Schizophrenia birth rates had several apparently non-random features in common with the SO1. The prominent SO1 fluctuation event that occurred between 1937 and 1943 is congruent with the most prominent fluctuation in schizophrenia birth rates. The relatively flat profile of SOI activity between 1927 and 1936 also corresponds to the flattest period in the schizophrenia time series. Both time series have prominent oscillations in the 3 ~, year range between 1946 and 1960. Significant associations between schizophrenia birth rates and measures of both sunshine and cloud cover were identified,and all three time series shared periodicity in the 3-4 year range. The analyses suggest that the risk of schizophrenia is higher for those born during times of increased cloud cover,reduced sunshine and positive SO1. These ecological analyses provide initial support for the vitamin D hypothesis, however alternative non-genetic candidate exposures also need to be considered. Other sites with year-to-year fluctuations in cloud cover and sunshine should examine patterns of association between these climate variables and schizophrenia birth rates. The Stanley Foundation supported this project.

The effect of solidification rate on the structure of magnesium-aluminium eutectic grains

A magnesium-aluminium alloy of eutectic composition was solidified under two different cooling conditions, producing a low and a high growth rate of the eutectic solid-liquid interface. The high growth rate specimen contained smaller eutectic grains and cells, with a smaller interphase spacing compared with the low growth rate specimen. The high growth rate specimen also contained some primary Mg17Al12 dendrites, suggesting that the coupled zone is skewed towards the Mg phase with increased undercooling, A lamellar eutectic morphology was observed in the low growth rate specimen, while the morphology was fibrous in the high growth rate specimen.

The interaction between the equity premium and the risk-free rate

The equity premium arises from the interaction between the atemporal risk premium for equity, the risk-free rate of intertemporal substitution and the impact of risk on the precautionary motive for saving. Depending on parameter values, the equity premium may either be increased or reduced by the presence of undiversifiable background risk. (C) 2000 Elsevier Science S.A. All rights reserved.

Exposure of rats to a high but not low dose of ethanol during early postnatal life increases the rate of loss of optic nerve axons and decreases the rate of myelination

Visual system abnormalities are commonly encountered in the fetal alcohol syndrome although the level of exposure at which they become manifest is uncertain. In this study we have examined the effects of either low (ETLD) or high dose (ETHD) ethanol, given between postnatal days 4-9, on the axons of the rat optic nerve. Rats were exposed to ethanol vapour in a special chamber for a period of 3 h per day during the treatment period. The blood alcohol concentration in the ETLD animals averaged similar to 171 mg/dl and in the ETHD animals similar to 430 mg/dl at the end of the treatment on any given day. Groups of 10 and 30-d-old mother-reared control (MRC), separation control (SC), ETLD and ETHD rats were anaesthetised with an intraperitoneal injection or ketamine and xylazine, and killed by intracardiac perfusion with phosphate-buffered glutaraldehyde. In the 10-d-old rat optic nerves there was a total of similar to 145000-165000 axons in MRC, SC and ETLD animals. About 4 % of these fibres were myelinated. The differences between these groups were not statistically significant. However, the 10-d-old ETHD animals had only about 75000 optic nerve axone (P < 0.05) of which about 2.8 % were myelinated. By 30 d of age there was a total of between 75000 90000 optic nerve axons, irrespective of the group examined. The proportion of axons which were myelinated at this age was still significantly lower (P < 0.001) in the ETHD animals (similar to 77 %) than in the other groups (about 98 %). It is concluded that the normal stages of development and maturation of the rat optic nerve axons, as assessed in this study, can be severely compromised by exposure to a relatively high (but not low) dose of ethanol between postnatal d 4 and 9.