Extreme water level exceedance probabilities around Australia

The potential impacts of extreme water level events on our coasts are increasing as populations grow and sea levels rise. To better prepare for the future, coastal engineers and managers need accurate estimates of average exceedance probabilities for extreme water levels. In this paper, we estimate present day probabilities of extreme water levels around the entire coastline of Australia. Tides and storm surges generated by extra-tropical storms were included by creating a 61-year (1949-2009) hindcast of water levels using a high resolution depth averaged hydrodynamic model driven with meteorological data from a global reanalysis. Tropical cyclone-induced surges were included through numerical modelling of a database of synthetic tropical cyclones equivalent to 10,000 years of cyclone activity around Australia. Predicted water level data was analysed using extreme value theory to construct return period curves for both the water level hindcast and synthetic tropical cyclone modelling. These return period curves were then combined by taking the highest water level at each return period.

Adaptive estimation of HMM transition probabilities

This paper presents new schemes for recursive estimation of the state transition probabilities for hidden Markov models (HMM's) via extended least squares (ELS) and recursive state prediction error (RSPE) methods. Local convergence analysis for the proposed RSPE algorithm is shown using the ordinary differential equation (ODE) approach developed for the more familiar recursive output prediction error (RPE) methods. The presented scheme converges and is relatively well conditioned compared with the ...

Adaptive estimation of hidden semi-Markov chains with parameterised transition probabilities and exponential decaying states

This paper develops maximum likelihood (ML) estimation schemes for finite-state semi-Markov chains in white Gaussian noise. We assume that the semi-Markov chain is characterised by transition probabilities of known parametric from with unknown parameters. We reformulate this hidden semi-Markov model (HSM) problem in the scalar case as a two-vector homogeneous hidden Markov model (HMM) problem in which the state consist of the signal augmented by the time to last transition. With this reformulation we apply the expectation Maximumisation (EM ) algorithm to obtain ML estimates of the transition probabilities parameters, Markov state levels and noise variance. To demonstrate our proposed schemes, motivated by neuro-biological applications, we use a damped sinusoidal parameterised function for the transition probabilities.

Kindergarten students' understanding of probability concepts

This study explored kindergarten students’ intuitive strategies and understandings in probabilities. The paper aims to provide an in depth insight into the levels of probability understanding across four constructs, as proposed by Jones (1997), for kindergarten students. Qualitative evidence from two students revealed that even before instruction pupils have a good capacity of predicting most and least likely events, of distinguishing fair probability situations from unfair ones, of comparing the probability of an event in two sample spaces, and of recognizing conditional probability events. These results contribute to the growing evidence on kindergarten students’ intuitive probabilistic reasoning. The potential of this study for improving the learning of probability, as well as suggestions for further research, are discussed.

Intelligent prognostics of machinery health utilising suspended condition monitoring data

The ability to forecast machinery failure is vital to reducing maintenance costs, operation downtime and safety hazards. Recent advances in condition monitoring technologies have given rise to a number of prognostic models for forecasting machinery health based on condition data. Although these models have aided the advancement of the discipline, they have made only a limited contribution to developing an effective machinery health prognostic system. The literature review indicates that there is not yet a prognostic model that directly models and fully utilises suspended condition histories (which are very common in practice since organisations rarely allow their assets to run to failure); that effectively integrates population characteristics into prognostics for longer-range prediction in a probabilistic sense; which deduces the non-linear relationship between measured condition data and actual asset health; and which involves minimal assumptions and requirements. This work presents a novel approach to addressing the above-mentioned challenges. The proposed model consists of a feed-forward neural network, the training targets of which are asset survival probabilities estimated using a variation of the Kaplan-Meier estimator and a degradation-based failure probability density estimator. The adapted Kaplan-Meier estimator is able to model the actual survival status of individual failed units and estimate the survival probability of individual suspended units. The degradation-based failure probability density estimator, on the other hand, extracts population characteristics and computes conditional reliability from available condition histories instead of from reliability data. The estimated survival probability and the relevant condition histories are respectively presented as “training target” and “training input” to the neural network. The trained network is capable of estimating the future survival curve of a unit when a series of condition indices are inputted. Although the concept proposed may be applied to the prognosis of various machine components, rolling element bearings were chosen as the research object because rolling element bearing failure is one of the foremost causes of machinery breakdowns. Computer simulated and industry case study data were used to compare the prognostic performance of the proposed model and four control models, namely: two feed-forward neural networks with the same training function and structure as the proposed model, but neglected suspended histories; a time series prediction recurrent neural network; and a traditional Weibull distribution model. The results support the assertion that the proposed model performs better than the other four models and that it produces adaptive prediction outputs with useful representation of survival probabilities. This work presents a compelling concept for non-parametric data-driven prognosis, and for utilising available asset condition information more fully and accurately. It demonstrates that machinery health can indeed be forecasted. The proposed prognostic technique, together with ongoing advances in sensors and data-fusion techniques, and increasingly comprehensive databases of asset condition data, holds the promise for increased asset availability, maintenance cost effectiveness, operational safety and – ultimately – organisation competitiveness.

Condition-based prognosis of machine health

Modern machines are complex and often required to operate long hours to achieve production targets. The ability to detect symptoms of failure, hence, forecasting the remaining useful life of the machine is vital to prevent catastrophic failures. This is essential to reducing maintenance cost, operation downtime and safety hazard. Recent advances in condition monitoring technologies have given rise to a number of prognosis models that attempt to forecast machinery health based on either condition data or reliability data. In practice, failure condition trending data are seldom kept by industries and data that ended with a suspension are sometimes treated as failure data. This paper presents a novel approach of incorporating historical failure data and suspended condition trending data in the prognostic model. The proposed model consists of a FFNN whose training targets are asset survival probabilities estimated using a variation of Kaplan-Meier estimator and degradation-based failure PDF estimator. The output survival probabilities collectively form an estimated survival curve. The viability of the model was tested using a set of industry vibration data.

Rover autonomous integrity monitoring of GNSS RTK positioning solutions with multi-constellations

This paper studies receiver autonomous integrity monitoring (RAIM) algorithms and performance benefits of RTK solutions with multiple-constellations. The proposed method is generally known as Multi-constellation RAIM -McRAIM. The McRAIM algorithms take advantage of the ambiguity invariant character to assist fast identification of multiple satellite faults in the context of multiple constellations, and then detect faulty satellites in the follow-up ambiguity search and position estimation processes. The concept of Virtual Galileo Constellation (VGC) is used to generate useful data sets of dual-constellations for performance analysis. Experimental results from a 24-h data set demonstrate that with GPS&VGC constellations, McRAIM can significantly enhance the detection and exclusion probabilities of two simultaneous faulty satellites in RTK solutions.

Identifying work related injuries : comparison of methods for interrogating text fields

Background: Work-related injuries in Australia are estimated to cost around $57.5 billion annually, however there are currently insufficient surveillance data available to support an evidence-based public health response. Emergency departments (ED) in Australia are a potential source of information on work-related injuries though most ED’s do not have an ‘Activity Code’ to identify work-related cases with information about the presenting problem recorded in a short free text field. This study compared methods for interrogating text fields for identifying work-related injuries presenting at emergency departments to inform approaches to surveillance of work-related injury.---------- Methods: Three approaches were used to interrogate an injury description text field to classify cases as work-related: keyword search, index search, and content analytic text mining. Sensitivity and specificity were examined by comparing cases flagged by each approach to cases coded with an Activity code during triage. Methods to improve the sensitivity and/or specificity of each approach were explored by adjusting the classification techniques within each broad approach.---------- Results: The basic keyword search detected 58% of cases (Specificity 0.99), an index search detected 62% of cases (Specificity 0.87), and the content analytic text mining (using adjusted probabilities) approach detected 77% of cases (Specificity 0.95).---------- Conclusions The findings of this study provide strong support for continued development of text searching methods to obtain information from routine emergency department data, to improve the capacity for comprehensive injury surveillance.

Micro-family-owned business : transitory survival mechanism or long term wealth creator?

Principal Topic: Entrepreneurship is key to employment, innovation and growth (Acs & Mueller, 2008), and as such, has been the subject of tremendous research in both the economic and management literatures since Solow (1957), Schumpeter (1934, 1943), and Penrose (1959). The presence of entrepreneurs in the economy is a key factor in the success or failure of countries to grow (Audretsch and Thurik, 2001; Dejardin, 2001). Further studies focus on the conditions of existence of entrepreneurship, influential factors invoked are historical, cultural, social, institutional, or purely economic (North, 1997; Thurik 1996 & 1999). Of particular interest, beyond the reasons behind the existence of entrepreneurship, are entrepreneurial survival and good ''performance'' factors. Using cross-country firm data analysis, La Porta & Schleifer (2008) confirm that informal micro-businesses provide on average half of all economic activity in developing countries. They find that these are utterly unproductive compared to formal firms, and conclude that the informal sector serves as a social security net ''keep[ing] millions of people alive, but disappearing over time'' (abstract). Robison (1986), Hill (1996, 1997) posit that the Indonesian government under Suharto always pointed to the lack of indigenous entrepreneurship , thereby motivating the nationalisation of all industries. Furthermore, the same literature also points to the fact that small businesses were mostly left out of development programmes because they were supposed less productive and having less productivity potential than larger ones. Vial (2008) challenges this view and shows that small firms represent about 70% of firms, 12% of total output, but contribute to 25% of total factor productivity growth on average over the period 1975-94 in the industrial sector (Table 10, p.316). ---------- Methodology/Key Propositions: A review of the empirical literature points at several under-researched questions. Firstly, we assess whether there is, evidence of small family-business entrepreneurship in Indonesia. Secondly, we examine and present the characteristics of these enterprises, along with the size of the sector, and its dynamics. Thirdly, we study whether these enterprises underperform compared to the larger scale industrial sector, as it is suggested in the literature. We reconsider performance measurements for micro-family owned businesses. We suggest that, beside productivity measures, performance could be appraised by both the survival probability of the firm, and by the amount of household assets formation. We compare micro-family-owned and larger industrial firms' survival probabilities after the 1997 crisis, their capital productivity, then compare household assets of families involved in business with those who do not. Finally, we examine human and social capital as moderators of enterprises' performance. In particular, we assess whether a higher level of education and community participation have an effect on the likelihood of running a family business, and whether it has an impact on households' assets level. We use the IFLS database compiled and published by RAND Corporation. The data is a rich community, households, and individuals panel dataset in four waves: 1993, 1997, 2000, 2007. We now focus on the waves 1997 and 2000 in order to investigate entrepreneurship behaviours in turbulent times, i.e. the 1997 Asian crisis. We use aggregate individual data, and focus on households data in order to study micro-family-owned businesses. IFLS data covers roughly 7,600 households in 1997 and over 10,000 households in 2000, with about 95% of 1997 households re-interviewed in 2000. Households were interviewed in 13 of the 27 provinces as defined before 2001. Those 13 provinces were targeted because accounting for 83% of the population. A full description of the data is provided in Frankenberg and Thomas (2000), and Strauss et alii (2004). We deflate all monetary values in Rupiah with the World Development Indicators Consumer Price Index base 100 in 2000. ---------- Results and Implications: We find that in Indonesia, entrepreneurship is widespread and two thirds of households hold one or several family businesses. In rural areas, in 2000, 75% of households run one or several businesses. The proportion of households holding both a farm and a non farm business is higher in rural areas, underlining the reliance of rural households on self-employment, especially after the crisis. Those businesses come in various sizes from very small to larger ones. The median business production value represents less than the annual national minimum wage. Figures show that at least 75% of farm businesses produce less than the annual minimum wage, with non farm businesses being more numerous to produce the minimum wage. However, this is only one part of the story, as production is not the only ''output'' or effect of the business. We show that the survival rate of those businesses ranks between 70 and 82% after the 1997 crisis, which contrasts with the 67% survival rate for the formal industrial sector (Ter Wengel & Rodriguez, 2006). Micro Family Owned Businesses might be relatively small in terms of production, they also provide stability in times of crisis. For those businesses that provide business assets figures, we show that capital productivity is fairly high, with rates that are ten times higher for non farm businesses. Results show that households running a business have larger family assets, and households are better off in urban areas. We run a panel logit model in order to test the effect of human and social capital on the existence of businesses among households. We find that non farm businesses are more likely to appear in households with higher human and social capital situated in urban areas. Farm businesses are more likely to appear in lower human capital and rural contexts, while still being supported by community participation. The estimation of our panel data model confirm that households are more likely to have higher family assets if situated in urban area, the higher the education level, the larger the assets, and running a business increase the likelihood of having larger assets. This is especially true for non farm businesses that have a clearly larger and more significant effect on assets than farm businesses. Finally, social capital in the form of community participation also has a positive effect on assets. Those results confirm the existence of a strong entrepreneurship culture among Indonesian households. Investigating survival rates also shows that those businesses are quite stable, even in the face of a violent crisis such as the 1997 one, and as a result, can provide a safety net. Finally, considering household assets - the returns of business to the household, rather than profit or productivity - the returns of business to itself, shows that households running a business are better off. While we demonstrate that uman and social capital are key to business existence, survival and performance, those results open avenues for further research regarding the factors that could hamper growth of those businesses in terms of output and employment.

Modelling Spikes in Electricity Prices

uring periods of market stress, electricity prices can rise dramatically. Electricity retailers cannot pass these extreme prices on to customers because of retail price regulation. Improved prediction of these price spikes therefore is important for risk management. This paper builds a time-varying-probability Markov-switching model of Queensland electricity prices, aimed particularly at forecasting price spikes. Variables capturing demand and weather patterns are used to drive the transition probabilities. Unlike traditional Markov-switching models that assume normality of the prices in each state, the model presented here uses a generalised beta distribution to allow for the skewness in the distribution of electricity prices during high-price episodes.

Optimising figure of merit for phonetic spoken term detection

This paper introduces a novel technique to directly optimise the Figure of Merit (FOM) for phonetic spoken term detection. The FOM is a popular measure of sTD accuracy, making it an ideal candiate for use as an objective function. A simple linear model is introduced to transform the phone log-posterior probabilities output by a phe classifier to produce enhanced log-posterior features that are more suitable for the STD task. Direct optimisation of the FOM is then performed by training the parameters of this model using a non-linear gradient descent algorithm. Substantial FOM improvements of 11% relative are achieved on held-out evaluation data, demonstrating the generalisability of the approach.

The application of phonetic distribution normalisation to likelihood-maximising speech enhancement for robust ASR

Traditional speech enhancement methods optimise signal-level criteria such as signal-to-noise ratio, but such approaches are sub-optimal for noise-robust speech recognition. Likelihood-maximising (LIMA) frameworks on the other hand, optimise the parameters of speech enhancement algorithms based on state sequences generated by a speech recogniser for utterances of known transcriptions. Previous applications of LIMA frameworks have generated a set of global enhancement parameters for all model states without taking in account the distribution of model occurrence, making optimisation susceptible to favouring frequently occurring models, in particular silence. In this paper, we demonstrate the existence of highly disproportionate phonetic distributions on two corpora with distinct speech tasks, and propose to normalise the influence of each phone based on a priori occurrence probabilities. Likelihood analysis and speech recognition experiments verify this approach for improving ASR performance in noisy environments.

Economics of maritime disasters : essays on the Titanic and Lusitania

This work seeks to fill some of the gap existing in the economics and behavioural economics literature pertaining to the decision making process of individuals under extreme environmental situations (life and death events). These essays specifically examine the sinking’s of the R.M.S. Titanic, on 14th April of 1912, and the R.M.S. Lusitania, on 7th May 1915, using econometric (multivariate) analysis techniques. The results show that even under extreme life and death conditions, social norms matter and are reflected in the survival probabilities of individuals onboard the Titanic. However, results from the comparative analysis of the Titanic and Lusitania show that social norms take time to organise and be effective. In the presence of such time constraints, the traditional “homo economicus” model of individual behaviour becomes evident as a survival of the fittest competition.

Reliability analysis with suspended items

Reliability analysis has several important engineering applications. Designers and operators of equipment are often interested in the probability of the equipment operating successfully to a given age - this probability is known as the equipment's reliability at that age. Reliability information is also important to those charged with maintaining an item of equipment, as it enables them to model and evaluate alternative maintenance policies for the equipment. In each case, information on failures and survivals of a typical sample of items is used to estimate the required probabilities as a function of the item's age, this process being one of many applications of the statistical techniques known as distribution fitting. In most engineering applications, the estimation procedure must deal with samples containing survivors (suspensions or censorings); this thesis focuses on several graphical estimation methods that are widely used for analysing such samples. Although these methods have been current for many years, they share a common shortcoming: none of them is continuously sensitive to changes in the ages of the suspensions, and we show that the resulting reliability estimates are therefore more pessimistic than necessary. We use a simple example to show that the existing graphical methods take no account of any service recorded by suspensions beyond their respective previous failures, and that this behaviour is inconsistent with one's intuitive expectations. In the course of this thesis, we demonstrate that the existing methods are only justified under restricted conditions. We present several improved methods and demonstrate that each of them overcomes the problem described above, while reducing to one of the existing methods where this is justified. Each of the improved methods thus provides a realistic set of reliability estimates for general (unrestricted) censored samples. Several related variations on these improved methods are also presented and justified. - i