Evaluating multivariate volatility forecasts : how effective are statistical and economic loss functions?

Multivariate volatility forecasts are an important input in many financial applications, in particular portfolio optimisation problems. Given the number of models available and the range of loss functions to discriminate between them, it is obvious that selecting the optimal forecasting model is challenging. The aim of this thesis is to thoroughly investigate how effective many commonly used statistical (MSE and QLIKE) and economic (portfolio variance and portfolio utility) loss functions are at discriminating between competing multivariate volatility forecasts. An analytical investigation of the loss functions is performed to determine whether they identify the correct forecast as the best forecast. This is followed by an extensive simulation study examines the ability of the loss functions to consistently rank forecasts, and their statistical power within tests of predictive ability. For the tests of predictive ability, the model confidence set (MCS) approach of Hansen, Lunde and Nason (2003, 2011) is employed. As well, an empirical study investigates whether simulation findings hold in a realistic setting. In light of these earlier studies, a major empirical study seeks to identify the set of superior multivariate volatility forecasting models from 43 models that use either daily squared returns or realised volatility to generate forecasts. This study also assesses how the choice of volatility proxy affects the ability of the statistical loss functions to discriminate between forecasts. Analysis of the loss functions shows that QLIKE, MSE and portfolio variance can discriminate between multivariate volatility forecasts, while portfolio utility cannot. An examination of the effective loss functions shows that they all can identify the correct forecast at a point in time, however, their ability to discriminate between competing forecasts does vary. That is, QLIKE is identified as the most effective loss function, followed by portfolio variance which is then followed by MSE. The major empirical analysis reports that the optimal set of multivariate volatility forecasting models includes forecasts generated from daily squared returns and realised volatility. Furthermore, it finds that the volatility proxy affects the statistical loss functions’ ability to discriminate between forecasts in tests of predictive ability. These findings deepen our understanding of how to choose between competing multivariate volatility forecasts.

Event-related brain potentials dissociate the developmental time-course of automatic numerical magnitude analysis and cognitive control functions during the first three years of primary school

In this study we set out to dissociate the developmental time course of automatic symbolic number processing and cognitive control functions in grade 1-3 British primary school children. Event-related potential (ERP) and behavioral data were collected in a physical size discrimination numerical Stroop task. Task-irrelevant numerical information was processed automatically already in grade 1. Weakening interference and strengthening facilitation indicated the parallel development of general cognitive control and automatic number processing. Relationships among ERP and behavioral effects suggest that control functions play a larger role in younger children and that automaticity of number processing increases from grade 1 to 3.

Structural damage detection using frequency response functions and neural networks

This paper illustrates the damage identification and condition assessment of a three story bookshelf structure using a new frequency response functions (FRFs) based damage index and Artificial Neural Networks (ANNs). A major obstacle of using measured frequency response function data is a large size input variables to ANNs. This problem is overcome by applying a data reduction technique called principal component analysis (PCA). In the proposed procedure, ANNs with their powerful pattern recognition and classification ability were used to extract damage information such as damage locations and severities from measured FRFs. Therefore, simple neural network models are developed, trained by Back Propagation (BP), to associate the FRFs with the damage or undamaged locations and severity of the damage of the structure. Finally, the effectiveness of the proposed method is illustrated and validated by using the real data provided by the Los Alamos National Laboratory, USA. The illustrated results show that the PCA based artificial Neural Network method is suitable and effective for damage identification and condition assessment of building structures. In addition, it is clearly demonstrated that the accuracy of proposed damage detection method can also be improved by increasing number of baseline datasets and number of principal components of the baseline dataset.

Structural damage identification with noise polluted Frequency Response Functions (FRFs)

Damage detection in structures has become increasingly important in recent years. While a number of damage detection and localization methods have been proposed, very few attempts have been made to explore the structure damage with noise polluted data which is unavoidable effect in real world. The measurement data are contaminated by noise because of test environment as well as electronic devices and this noise tend to give error results with structural damage identification methods. Therefore it is important to investigate a method which can perform better with noise polluted data. This paper introduces a new damage index using principal component analysis (PCA) for damage detection of building structures being able to accept noise polluted frequency response functions (FRFs) as input. The FRF data are obtained from the function datagen of MATLAB program which is available on the web site of the IASC-ASCE (International Association for Structural Control– American Society of Civil Engineers) Structural Health Monitoring (SHM) Task Group. The proposed method involves a five-stage process: calculation of FRFs, calculation of damage index values using proposed algorithm, development of the artificial neural networks and introducing damage indices as input parameters and damage detection of the structure. This paper briefly describes the methodology and the results obtained in detecting damage in all six cases of the benchmark study with different noise levels. The proposed method is applied to a benchmark problem sponsored by the IASC-ASCE Task Group on Structural Health Monitoring, which was developed in order to facilitate the comparison of various damage identification methods. The illustrated results show that the PCA-based algorithm is effective for structural health monitoring with noise polluted FRFs which is of common occurrence when dealing with industrial structures.

IL-18 inhibits growth of murine orthotopic prostate carcinomas via both adaptive and innate immune mechanisms

Interleukin(IL)-18 is a pleiotrophic cytokine with functions in immune modulation, angiogenesis and bone metabolism. In this study, the potential of IL-18 as an immunotherapy for prostate cancer (PCa) was examined using the murine model of prostate carcinoma, RM1 and a bone metastatic variant RM1(BM)/B4H7-luc. RM1 and RM1(BM)/B4H7-luc cells were stably transfected to express bioactive IL-18. These cells were implanted into syngeneic immunocompetent mice, with or without an IL-18-neutralising antibody (αIL-18, SK113AE4). IL-18 significantly inhibited the growth of both subcutaneous and orthotopic RM1 tumors and the IL-18 neutralizing antibody abrogated the tumor growth-inhibition. In vivo neutralization of interferon-gamma (IFN-γ) completely eliminated the anti-tumor effects of IL-18 confirming an essential role of IFN-γ as a down-stream mediator of the anti-tumor activity of IL-18. Tumors from mice in which IL-18 and/or IFN-γ was neutralized contained significantly fewer CD4+ and CD8+ T cells than those with functional IL-18. The essential role of adaptive immunity was demonstrated as tumors grew more rapidly in RAG1−/− mice or in mice depleted of CD4+ and/or CD8+ cells than in normal mice. The tumors in RAG1−/− mice were also significantly smaller when IL-18 was present, indicating that innate immune mechanisms are involved. IL-18 also induced an increase in tumor infiltration of macrophages and neutrophils but not NK cells. In other experiments, direct injection of recombinant IL-18 into established tumors also inhibited tumor growth, which was associated with an increase in intratumoral macrophages, but not T cells. These results suggest that local IL-18 in the tumor environment can significantly potentiate anti-tumor immunity in the prostate and clearly demonstrate that this effect is mediated by innate and adaptive immune mechanisms.

Cytokine synthesis by intestinal intraepithelial lymphocytes. Both gamma/delta T cell receptor-positive and alpha/beta T cell receptor-positive T cells in the G1 phase of cell cycle produce IFN-gamma and IL-5

Murine intestinal intraepithelial lymphocytes (IEL) have been shown to contain subsets of alpha/beta TCR+ and gamma/delta TCR+ T cells that spontaneously produce cytokines such as IFN-gamma and IL-5. We have now determined the nature and cell cycle stage of these cytokine-producing T lymphocytes in EIL by using IFN-gamma- and IL-5-specific ELISPOT assay, cytokine-specific mRNA-cDNA dot-blot hybridization and polymerase chain reaction, and flow cytometry (FACS) for DNA analysis. When CD3+ T cells from IEL of normal C3H/HeN mice were separated into low and high density fractions by discontinuous Percoll gradients, IFN-gamma and IL-5 spot-forming cells were only found in the former population. Analysis of mRNA for these cytokines by both IFN-gamma- and IL-5-specific dot-blot hybridization and polymerase chain reaction revealed that higher levels of message for IFN-gamma and IL-5 were also seen in the low density fraction. However, cell cycle analysis of these two fractions by FACS using propidium iodide showed a similar pattern of cell cycle stages in both low and high density populations (G0 + G1 approximately 96 to 98% and S/G2 + M approximately 2 to 4%). Finally, mRNA from gamma/delta TCR+ and alpha/beta TCR+ T cells in both low and high density fractions of IEL were analyzed for IFN-gamma and IL-5 message by polymerase chain reaction. After 35 cycles of amplification, both gamma/delta TCR+ and alpha/beta TCR+ T cells in the low density fraction expressed higher levels of message for these two cytokines when compared with the high density population. These results have now shown that both gamma/delta and alpha/beta TCR+ IEL can be separated into low and high density subsets and both fractions possess a similar stage of cell cycle. However, only the low density cells (in G1 phase) of both gamma/delta and alpha/beta TCR types possess increased cytokine-specific mRNA and produce the cytokines IFN-gamma and IL-5. Our results suggest that alpha/beta TCR+ and gamma/delta TCR+ IEL can produce cytokines without cell proliferation.

A comparison of safety climate and safety outcomes between construction and resource functions in a large case study organisation

The research described in this paper forms part of an in-depth investigation of safety culture in one of Australia’s largest construction companies. The research builds on a previous qualitative study with organisational safety leaders and further investigates how safety culture is perceived and experienced by organisational members, as well as how this relates to their safety behaviour and related outcomes at work. Participants were 2273 employees of the case study organisation, with 689 from the Construction function and 1584 from the Resources function. The results of several analyses revealed some interesting organisational variance on key measures. Specifically, the Construction function scored significantly higher on all key measures: safety climate, safety motivation, safety compliance, and safety participation. The results are discussed in terms of relevance in an applied research context.

Planar difference functions

In 1980 Alltop produced a family of cubic phase sequences that nearly meet the Welch bound for maximum non-peak correlation magnitude. This family of sequences were shown by Wooters and Fields to be useful for quantum state tomography. Alltop’s construction used a function that is not planar, but whose difference function is planar. In this paper we show that Alltop type functions cannot exist in fields of characteristic 3 and that for a known class of planar functions, x^3 is the only Alltop type function.

Principles of motor learning in ecological dynamics : a comment on functions of learning and the acquisition of motor skills (with reference to sport)

This paper provides a commentary on the contribution by Dr Chow who questioned whether the functions of learning are general across all categories of tasks or whether there are some task-particular aspects to the functions of learning in relation to task type. Specifically, they queried whether principles and practice for the acquisition of sport skills are different than what they are for musical, industrial, military and human factors skills. In this commentary we argue that ecological dynamics contains general principles of motor learning that can be instantiated in specific performance contexts to underpin learning design. In this proposal, we highlight the importance of conducting skill acquisition research in sport, rather than relying on empirical outcomes of research from a variety of different performance contexts. Here we discuss how task constraints of different performance contexts (sport, industry, military, music) provide different specific information sources that individuals use to couple their actions when performing and acquiring skills. We conclude by suggesting that his relationship between performance task constraints and learning processes might help explain the traditional emphasis on performance curves and performance outcomes to infer motor learning.