Exact and approximate Bayesian inference for low count time series models with intractable likelihoods

In this paper we present a new simulation methodology in order to obtain exact or approximate Bayesian inference for models for low-valued count time series data that have computationally demanding likelihood functions. The algorithm fits within the framework of particle Markov chain Monte Carlo (PMCMC) methods. The particle filter requires only model simulations and, in this regard, our approach has connections with approximate Bayesian computation (ABC). However, an advantage of using the PMCMC approach in this setting is that simulated data can be matched with data observed one-at-a-time, rather than attempting to match on the full dataset simultaneously or on a low-dimensional non-sufficient summary statistic, which is common practice in ABC. For low-valued count time series data we find that it is often computationally feasible to match simulated data with observed data exactly. Our particle filter maintains $N$ particles by repeating the simulation until $N+1$ exact matches are obtained. Our algorithm creates an unbiased estimate of the likelihood, resulting in exact posterior inferences when included in an MCMC algorithm. In cases where exact matching is computationally prohibitive, a tolerance is introduced as per ABC. A novel aspect of our approach is that we introduce auxiliary variables into our particle filter so that partially observed and/or non-Markovian models can be accommodated. We demonstrate that Bayesian model choice problems can be easily handled in this framework.

Comparison between analytical and 3D finite element solutions for borehole stresses in anisotropic elastic rock

We present a rigorous validation of the analytical Amadei solution for the stress concentration around an arbitrarily orientated borehole in general anisotropic elastic media. First, we revisit the theoretical framework of the Amadei solution and present analytical insights that show that the solution does indeed contain all special cases of symmetry, contrary to previous understanding, provided that the reduced strain coefficients b11 and b55 are not equal. It is shown from theoretical considerations and published experimental data that the b11 and b55 are not equal for realistic rocks. Second, we develop a 3D finite element elastic model within a hybrid analytical–numerical workflow that circumvents the need to rebuild and remesh the model for every borehole and material orientation. Third, we show that the borehole stresses computed from the numerical model and the analytical solution match almost perfectly for different borehole orientations (vertical, deviated and horizontal) and for several cases involving isotropic, transverse isotropic and orthorhombic symmetries. It is concluded that the analytical Amadei solution is valid with no restriction on the borehole orientation or the symmetry of the elastic anisotropy.

Long exposure localization in darkness using consumer cameras

In this paper we demonstrate passive vision-based localization in environments more than two orders of magnitude darker than the current benchmark using a 100 webcam and a 500 camera. Our approach uses the camera’s maximum exposure duration and sensor gain to achieve appropriately exposed images even in unlit night-time environments, albeit with extreme levels of motion blur. Using the SeqSLAM algorithm, we first evaluate the effect of variable motion blur caused by simulated exposures of 132 ms to 10000 ms duration on localization performance. We then use actual long exposure camera datasets to demonstrate day-night localization in two different environments. Finally we perform a statistical analysis that compares the baseline performance of matching unprocessed greyscale images to using patch normalization and local neighbourhood normalization – the two key SeqSLAM components. Our results and analysis show for the first time why the SeqSLAM algorithm is effective, and demonstrate the potential for cheap camera-based localization systems that function across extreme perceptual change.

Predicate encryption with various properties

Predicate encryption (PE) is a new primitive which supports exible control over access to encrypted data. In PE schemes, users' decryption keys are associated with predicates f and ciphertexts encode attributes a that are specified during the encryption procedure. A user can successfully decrypt if and only if f(a) = 1. In this thesis, we will investigate several properties that are crucial to PE. We focus on expressiveness of PE, Revocable PE and Hierarchical PE (HPE) with forward security. For all proposed systems, we provide a security model and analysis using the widely accepted computational complexity approach. Our first contribution is to explore the expressiveness of PE. Existing PE supports a wide class of predicates such as conjunctions of equality, comparison and subset queries, disjunctions of equality queries, and more generally, arbitrary combinations of conjunctive and disjunctive equality queries. We advance PE to evaluate more expressive predicates, e.g., disjunctive comparison or disjunctive subset queries. Such expressiveness is achieved at the cost of computational and space overhead. To improve the performance, we appropriately revise the PE to reduce the computational and space cost. Furthermore, we propose a heuristic method to reduce disjunctions in the predicates. Our schemes are proved in the standard model. We then introduce the concept of Revocable Predicate Encryption (RPE), which extends the previous PE setting with revocation support: private keys can be used to decrypt an RPE ciphertext only if they match the decryption policy (defined via attributes encoded into the ciphertext and predicates associated with private keys) and were not revoked by the time the ciphertext was created. We propose two RPE schemes. Our first scheme, termed Attribute- Hiding RPE (AH-RPE), offers attribute-hiding, which is the standard PE property. Our second scheme, termed Full-Hiding RPE (FH-RPE), offers even stronger privacy guarantees, i.e., apart from possessing the Attribute-Hiding property, the scheme also ensures that no information about revoked users is leaked from a given ciphertext. The proposed schemes are also proved to be secure under well established assumptions in the standard model. Secrecy of decryption keys is an important pre-requisite for security of (H)PE and compromised private keys must be immediately replaced. The notion of Forward Security (FS) reduces damage from compromised keys by guaranteeing confidentiality of messages that were encrypted prior to the compromise event. We present the first Forward-Secure Hierarchical Predicate Encryption (FS-HPE) that is proved secure in the standard model. Our FS-HPE scheme offers some desirable properties: time-independent delegation of predicates (to support dynamic behavior for delegation of decrypting rights to new users), local update for users' private keys (i.e., no master authority needs to be contacted), forward security, and the scheme's encryption process does not require knowledge of predicates at any level including when those predicates join the hierarchy.

Non-contact ACL injuries : association with clegg soil impact test values for sports fields, soil moisture and prevailing weather

Introduction: There is a recognised relationship between dry weather conditions and increased risk of anterior cruciate ligament (ACL) injury. Previous studies have identified 28 day evaporation as an important weather-based predictor of non-contact ACL injuries in professional Australian Football League matches. The mechanism of non-contact injury to the ACL is believed to increased traction and impact forces between footwear and playing surface. Ground hardness and the amount and quality of grass are factors that would most likely influence this and are inturn, related to the soil moisture content and prevailing weather conditions. This paper explores the relationship between soil moisture content, preceding weather conditions and the Clegg Soil Impact Test (CSIT) which is an internationally recognised standard measure of ground hardness for sports fields. Methodology: The 2.25 kg Clegg Soil Impact Test and a pair of 12 cm soil moisture probes were used to measure ground hardness and percentage moisture content. Five football fields were surveyed at 13 prescribed sites just before seven football matches from October 2008 to January 2009 (an FC Women’s WLeague team). Weather conditions recorded at the nearest weather station were obtained from the Bureau of Meteorology website and total rainfall less evaporation was calculated for 7 and 28 days prior to each match. All non-contact injuries occurring during match play and their location on the field were recorded. Results/conclusions: Ground hardness varied between CSIT 5 and 17 (x10G) (8 is considered a good value for sports fields). Variations within fields were typically greatest in the centre and goal areas. Soil moisture ranged from 3 to 40% with some fields requiring twice the moisture content of others to maintain similar CSIT values. There was a non-linear, negative relationship for ground hardness versus moisture content and a linear relationship with weather (R2, of 0.30 and 0.34, respectively). Three non-contact ACL injuries occurred during the season. Two of these were associated with hard and variable ground conditions.

Mathematical modelling of LiFePO4 cathodes

LiFePO4 is a commercially available battery material with good theoretical discharge capacity, excellent cycle life and increased safety compared with competing Li-ion chemistries. It has been the focus of considerable experimental and theoretical scrutiny in the past decade, resulting in LiFePO4 cathodes that perform well at high discharge rates. This scrutiny has raised several questions about the behaviour of LiFePO4 material during charge and discharge. In contrast to many other battery chemistries that intercalate homogeneously, LiFePO4 can phase-separate into highly and lowly lithiated phases, with intercalation proceeding by advancing an interface between these two phases. The main objective of this thesis is to construct mathematical models of LiFePO4 cathodes that can be validated against experimental discharge curves. This is in an attempt to understand some of the multi-scale dynamics of LiFePO4 cathodes that can be difficult to determine experimentally. The first section of this thesis constructs a three-scale mathematical model of LiFePO4 cathodes that uses a simple Stefan problem (which has been used previously in the literature) to describe the assumed phase-change. LiFePO4 crystals have been observed agglomerating in cathodes to form a porous collection of crystals and this morphology motivates the use of three size-scales in the model. The multi-scale model developed validates well against experimental data and this validated model is then used to examine the role of manufacturing parameters (including the agglomerate radius) on battery performance. The remainder of the thesis is concerned with investigating phase-field models as a replacement for the aforementioned Stefan problem. Phase-field models have recently been used in LiFePO4 and are a far more accurate representation of experimentally observed crystal-scale behaviour. They are based around the Cahn-Hilliard-reaction (CHR) IBVP, a fourth-order PDE with electrochemical (flux) boundary conditions that is very stiff and possesses multiple time and space scales. Numerical solutions to the CHR IBVP can be difficult to compute and hence a least-squares based Finite Volume Method (FVM) is developed for discretising both the full CHR IBVP and the more traditional Cahn-Hilliard IBVP. Phase-field models are subject to two main physicality constraints and the numerical scheme presented performs well under these constraints. This least-squares based FVM is then used to simulate the discharge of individual crystals of LiFePO4 in two dimensions. This discharge is subject to isotropic Li+ diffusion, based on experimental evidence that suggests the normally orthotropic transport of Li+ in LiFePO4 may become more isotropic in the presence of lattice defects. Numerical investigation shows that two-dimensional Li+ transport results in crystals that phase-separate, even at very high discharge rates. This is very different from results shown in the literature, where phase-separation in LiFePO4 crystals is suppressed during discharge with orthotropic Li+ transport. Finally, the three-scale cathodic model used at the beginning of the thesis is modified to simulate modern, high-rate LiFePO4 cathodes. High-rate cathodes typically do not contain (large) agglomerates and therefore a two-scale model is developed. The Stefan problem used previously is also replaced with the phase-field models examined in earlier chapters. The results from this model are then compared with experimental data and fit poorly, though a significant parameter regime could not be investigated numerically. Many-particle effects however, are evident in the simulated discharges, which match the conclusions of recent literature. These effects result in crystals that are subject to local currents very different from the discharge rate applied to the cathode, which impacts the phase-separating behaviour of the crystals and raises questions about the validity of using cathodic-scale experimental measurements in order to determine crystal-scale behaviour.

Iris based identity verification robust to sample presentation security attacks

Iris based identity verification is highly reliable but it can also be subject to attacks. Pupil dilation or constriction stimulated by the application of drugs are examples of sample presentation security attacks which can lead to higher false rejection rates. Suspects on a watch list can potentially circumvent the iris based system using such methods. This paper investigates a new approach using multiple parts of the iris (instances) and multiple iris samples in a sequential decision fusion framework that can yield robust performance. Results are presented and compared with the standard full iris based approach for a number of iris degradations. An advantage of the proposed fusion scheme is that the trade-off between detection errors can be controlled by setting parameters such as the number of instances and the number of samples used in the system. The system can then be operated to match security threat levels. It is shown that for optimal values of these parameters, the fused system also has a lower total error rate.

The effectiveness of public sector asset management in Malaysia

Property is an important factor in all businesses production in order to function. Nourse (1990) quoted ¡°some businesses are real estate, all businesses use real estate¡±. In recent years, the management of property assets has become the focus of many organisations, including the non-real estate businesses. Good asset management is concerned with the effective utilisation of a property owner.s assets. It is the management process of ensuring that the portfolio of properties held meets the overall requirements of the users. In short, it is the process of identifying the user.s requirement and the rationalisation of property holdings to match that requirement best, followed by a monitoring and ongoing review of practice. In Malaysia, federal agencies and local authorities are among the largest property asset owners. Recently the federal government has released a Total Asset Management Manual (TAMM). It is at the preliminary stage of implementation. This thesis will study the international practices of asset management of public sector assets and assess the effectiveness of TAMM. This research will focus on current international practices for the effective management of public sector property assets. The current application in Malaysia will be highlighted, to determine the awareness and understanding of the current practices to the recently released TAMM. This research is an exploratory research. The basis of this research relies on the combination of qualitative and quantitative approach, whereby the qualitative approach focuses on the international practices and its application to the management of public sector property assets. Questionnaires survey will be conducted among the Malaysian public property assets managers and users in the quantitative approach to gauge the collective opinion on the current practices of TAMM and its implementation

Molecular dynamics simulation of structural characteristics in metal cluster deposition on surfaces

The deposition of small metal clusters (Cu, Au and Al) on f.c.c. metals (Cu, Au and Ni) has been studied by molecular dynamics simulation using Finnis–Sinclair (FS) potential. The impact energy varied from 0.01 to 10 eV/atom. First, the deposition of single cluster was simulated. We observed that, even at much lower energy, a small cluster with (Ih) icosahedral symmetry was reconstructed to match the substrate structure (f.c.c.) after deposition. Next, clusters were modeled to drop, one after the other, on the surface. The nanostructure was found by soft landing of Au clusters on Cu with increasing coverage, where interfacial energy dominates. While at relatively higher deposition energy (a few eV), the ordered f.c.c.-like structure was observed in the first adlayer of the film formed by Al clusters depositing on Ni substrate. This characteristic is mainly attributive to the ballistic collision. Our results indicate that the surface morphology synthesized by cluster deposition could be controlled by experimental parameters, which will be helpful for controlled design of nanostructure.

Vision-based place recognition : how low can you go?

In this paper we use the algorithm SeqSLAM to address the question, how little and what quality of visual information is needed to localize along a familiar route? We conduct a comprehensive investigation of place recognition performance on seven datasets while varying image resolution (primarily 1 to 512 pixel images), pixel bit depth, field of view, motion blur, image compression and matching sequence length. Results confirm that place recognition using single images or short image sequences is poor, but improves to match or exceed current benchmarks as the matching sequence length increases. We then present place recognition results from two experiments where low-quality imagery is directly caused by sensor limitations; in one, place recognition is achieved along an unlit mountain road by using noisy, long-exposure blurred images, and in the other, two single pixel light sensors are used to localize in an indoor environment. We also show failure modes caused by pose variance and sequence aliasing, and discuss ways in which they may be overcome. By showing how place recognition along a route is feasible even with severely degraded image sequences, we hope to provoke a re-examination of how we develop and test future localization and mapping systems.

Gelatin-methacrylamide hydrogels as potential biomaterials for fabrication of tissue-engineered cartilage constructs

Gelatin-methacrylamide (gelMA) hydrogels are shown to support chondrocyte viability and differentiation and give wide ranging mechanical properties depending on several cross-linking parameters. Polymer concentration, UV exposure time, and thermal gelation prior to UV exposure allow for control over hydrogel stiffness and swelling properties. GelMA solutions have a low viscosity at 37 °C, which is incompatible with most biofabrication approaches. However, incorporation of hyaluronic acid (HA) and/or co-deposition with thermoplastics allows gelMA to be used in biofabrication processes. These attributes may allow engineered constructs to match the natural functional variations in cartilage mechanical and geometrical properties.

Epistatic effects of potassium channel variation on cardiac repolarization and atrial fibrillation risk

Objectives The aim of this study was to evaluate the role of cardiac K+ channel gene variants in families with atrial fibrillation (AF). Background The K+ channels play a major role in atrial repolarization but single mutations in cardiac K+ channel genes are infrequently present in AF families. The collective effect of background K+ channel variants of varying prevalence and effect size on the atrial substrate for AF is largely unexplored. Methods Genes encoding the major cardiac K+ channels were resequenced in 80 AF probands. Nonsynonymous coding sequence variants identified in AF probands were evaluated in 240 control subjects. Novel variants were characterized using patch-clamp techniques and in silico modeling was performed using the Courtemanche atrial cell model. Results Nineteen nonsynonymous variants in 9 genes were found, including 11 rare variants. Rare variants were more frequent in AF probands (18.8% vs. 4.2%, p < 0.001), and the mean number of variants was greater (0.21 vs. 0.04, p < 0.001). The majority of K+ channel variants individually had modest functional effects. Modeling simulations to evaluate combinations of K+ channel variants of varying population frequency indicated that simultaneous small perturbations of multiple current densities had nonlinear interactions and could result in substantial (>30 ms) shortening or lengthening of action potential duration as well as increased dispersion of repolarization. Conclusions Families with AF show an excess of rare functional K+ channel gene variants of varying phenotypic effect size that may contribute to an atrial arrhythmogenic substrate. Atrial cell modeling is a useful tool to assess epistatic interactions between multiple variants.

Association study of calcitonin gene-related polypeptide-alpha (CALCA) gene polymorphism with migraine

Migraine is a neurological disorder that is associated with increased levels of calcitonin gene-related peptide (CGRP) in plasma. CGRP, being one of the mediators of neurogenic inflammation and a phenomenon implicated in the pathogenesis of migraine headache, is thus suggested to have an important role in migraine pathophysiology. Polymorphisms of the CALCA gene have been linked to Parkinson's disease, ovarian cancer and essential hypertension, suggesting a functional role for these polymorphisms. Given the strong evidence linking CGRP and migraine, it is hypothesised that polymorphisms in the CALCA gene may play a role in migraine pathogenesis. Seemingly non functional intronic polymorphisms are capable of disrupting normal RNA processing or introducing a splice site in the transcript. A 16 bp deletion in the first intron of the CALCA gene has been reported to be a good match for the binding site for a transcription factor expressed strongly in neural crest derived cells, AP-2. This deletion also eliminates an intron splicing enhancer (ISE) that may potentially cause exon skipping. This study investigated the role of the 16 bp intronic deletion in the CALCA gene in migraineurs and matched control individuals. Six hundred individuals were genotyped for the deletion by polymerase chain reaction followed by fragment analysis on the 3130 Genetic Analyser. The results of this study showed no significant association between the intronic 16 bp deletion in the CALCA gene and migraine in the tested Australian Caucasian population. However, given the evidence linking CGRP and migraine, further investigation of variants with this gene may be warranted.

Histogram of weighted local directions for gait recognition

In this paper, we explore the effectiveness of patch-based gradient feature extraction methods when applied to appearance-based gait recognition. Extending existing popular feature extraction methods such as HOG and LDP, we propose a novel technique which we term the Histogram of Weighted Local Directions (HWLD). These 3 methods are applied to gait recognition using the GEI feature, with classification performed using SRC. Evaluations on the CASIA and OULP datasets show significant improvements using these patch-based methods over existing implementations, with the proposed method achieving the highest recognition rate for the respective datasets. In addition, the HWLD can easily be extended to 3D, which we demonstrate using the GEV feature on the DGD dataset, observing improvements in performance.

Integrating collaborative filtering and matching-based search for product recommendations

Currently, recommender systems (RS) have been widely applied in many commercial e-commerce sites to help users deal with the information overload problem. Recommender systems provide personalized recommendations to users and, thus, help in making good decisions about which product to buy from the vast amount of product choices. Many of the current recommender systems are developed for simple and frequently purchased products like books and videos, by using collaborative-filtering and content-based approaches. These approaches are not directly applicable for recommending infrequently purchased products such as cars and houses as it is difficult to collect a large number of ratings data from users for such products. Many of the ecommerce sites for infrequently purchased products are still using basic search-based techniques whereby the products that match with the attributes given in the target user’s query are retrieved and recommended. However, search-based recommenders cannot provide personalized recommendations. For different users, the recommendations will be the same if they provide the same query regardless of any difference in their interest. In this article, a simple user profiling approach is proposed to generate user’s preferences to product attributes (i.e., user profiles) based on user product click stream data. The user profiles can be used to find similarminded users (i.e., neighbours) accurately. Two recommendation approaches are proposed, namely Round- Robin fusion algorithm (CFRRobin) and Collaborative Filtering-based Aggregated Query algorithm (CFAgQuery), to generate personalized recommendations based on the user profiles. Instead of using the target user’s query to search for products as normal search based systems do, the CFRRobin technique uses the attributes of the products in which the target user’s neighbours have shown interest as queries to retrieve relevant products, and then recommends to the target user a list of products by merging and ranking the returned products using the Round Robin method. The CFAgQuery technique uses the attributes of the products that the user’s neighbours have shown interest in to derive an aggregated query, which is then used to retrieve products to recommend to the target user. Experiments conducted on a real e-commerce dataset show that both the proposed techniques CFRRobin and CFAgQuery perform better than the standard Collaborative Filtering and the Basic Search approaches, which are widely applied by the current e-commerce applications.