Modelling nonlinear stochastic dynamics in financial time series

For analysing financial time series two main opposing viewpoints exist, either capital markets are completely stochastic and therefore prices follow a random walk, or they are deterministic and consequently predictable. For each of these views a great variety of tools exist with which it can be tried to confirm the hypotheses. Unfortunately, these methods are not well suited for dealing with data characterised in part by both paradigms. This thesis investigates these two approaches in order to model the behaviour of financial time series. In the deterministic framework methods are used to characterise the dimensionality of embedded financial data. The stochastic approach includes here an estimation of the unconditioned and conditional return distributions using parametric, non- and semi-parametric density estimation techniques. Finally, it will be shown how elements from these two approaches could be combined to achieve a more realistic model for financial time series.

HRM systems, internal marketing and performance in Indian call centres:non-technical summary (Research summary) ESRC end of award report, RES-000-22-1876. Swindon: ESRC

Considering the rapid growth of call centres (CCs) in India, its implications for businesses in the UK and a scarcity of research on human resource management (HRM) related issues in Indian CCs, this research has two main aims. First, to highlight the nature of HRM systems relevant to Indian call centres. Second, to understand the significance of internal marketing (IM) in influencing the frontline employees’ job-related attitudes and performance. Rewards being an important component of IM, the relationships between different types of rewards as part of an IM strategy, attitudes and performance of employees in Indian CCs will also be examined. Further, the research will investigate which type of commitment mediates the link between rewards and performance and why. The data collection will be via two phases. The first phase would involve a series of in-depth interviews with both the managers and employees to understand the functioning of CCs, and development of suitable HRM systems for the Indian context. The second phase would involve data collection through questionnaires distributed to the frontline employees and supervisors to examine the relationships among IM, employee attitudes and performance. Such an investigation is expected to contribute to development of better theory and practice.

Une façon d’indiquer la «non-coïncidence entre les mots et les choses», on va dire…

Nous souhaitons nous pencher ici sur un emploi particulier de la périphrase en aller + infinitif qui n’a fait l’objet – à notre connaissance – que d’un article (Lansari 2010). Cet emploi « modalisant » que Lansari limite à la formule 'on va dire' mériterait d’être approfondi pour plusieurs raisons. D’une part, l’emploi n’est décrit que sur base de « vingt exemples tirés d’internet, de blogs ou de forums » (Lansari 2010: 120) alors que, de l’aveu de Lansari elle-même, l’emploi relève de l’oral. Il serait donc utile d’enrichir – quantitativement et qualitativement - le corpus et d’y intégrer des occurrences d’oral authentique. D’autre part, Lansari restreint l’emploi modalisant à la séquence 'on va dire' ; on pourrait s’interroger sur la capacité de séquences comme 'je vais dire' à remplir les mêmes fonctions discursives. Dans cet article, nous commencerons par un – forcément bref – état de la question. Après avoir présenté le corpus, nous testerons les hypothèses précédemment défendues à la lueur du corpus rassemblé: (a) Le corpus CFPP2000 issu du projet Discours sur la ville. Corpus de Français Parlé Parisien des années 2000 (disponible en ligne à http://cfpp2000.univ-paris3.fr/Corpus.html). CFPP2000 donne la parole à 41 informateurs en 28 interviews (2198 min) et a généré 96 occurrences de on va dire modalisant. (b) Le corpus CLAPI comprenant 45 heures d’interactions interrogeables en ligne à http://clapi.univ-lyon2.fr/analyse_requete_aide.php?menu=outils. On y a relevé 12 exemples de on va dire modalisant. (c) Un corpus personnel d’interviews (163min) réalisées pendant l’année académique 2009-10 auprès de cinq étudiants Erasmus français grâce au soutien d’une bourse de la Délégation Générale à la Langue Française et aux Langues de France (DGLFLF). Les entretiens avec une assistante de recherche, basés sur les thèmes suivants, étaient supposés générer l’emploi d’une variété de temps verbaux : - Récits de rêve (imparfait) - Récits biographiques (personnage historique vs autobiographie) (PC vs PS) - Narration de film vs d’épisode historique (PC/ PRES vs PS) - Présentation de projets d’avenir vs conjectures (Futur périphrastique ou simple) Le corpus contient dix-sept occurrences de on va dire générés par deux des cinq informateurs : 15 par A. et 2 par J. Notre réflexion se basera donc sur 125 occurrences orales de 'on va dire'.

Non-linear hierarchical visualisation

This thesis applies a hierarchical latent trait model system to a large quantity of data. The motivation for it was lack of viable approaches to analyse High Throughput Screening datasets which maybe include thousands of data points with high dimensions. High Throughput Screening (HTS) is an important tool in the pharmaceutical industry for discovering leads which can be optimised and further developed into candidate drugs. Since the development of new robotic technologies, the ability to test the activities of compounds has considerably increased in recent years. Traditional methods, looking at tables and graphical plots for analysing relationships between measured activities and the structure of compounds, have not been feasible when facing a large HTS dataset. Instead, data visualisation provides a method for analysing such large datasets, especially with high dimensions. So far, a few visualisation techniques for drug design have been developed, but most of them just cope with several properties of compounds at one time. We believe that a latent variable model (LTM) with a non-linear mapping from the latent space to the data space is a preferred choice for visualising a complex high-dimensional data set. As a type of latent variable model, the latent trait model can deal with either continuous data or discrete data, which makes it particularly useful in this domain. In addition, with the aid of differential geometry, we can imagine the distribution of data from magnification factor and curvature plots. Rather than obtaining the useful information just from a single plot, a hierarchical LTM arranges a set of LTMs and their corresponding plots in a tree structure. We model the whole data set with a LTM at the top level, which is broken down into clusters at deeper levels of t.he hierarchy. In this manner, the refined visualisation plots can be displayed in deeper levels and sub-clusters may be found. Hierarchy of LTMs is trained using expectation-maximisation (EM) algorithm to maximise its likelihood with respect to the data sample. Training proceeds interactively in a recursive fashion (top-down). The user subjectively identifies interesting regions on the visualisation plot that they would like to model in a greater detail. At each stage of hierarchical LTM construction, the EM algorithm alternates between the E- and M-step. Another problem that can occur when visualising a large data set is that there may be significant overlaps of data clusters. It is very difficult for the user to judge where centres of regions of interest should be put. We address this problem by employing the minimum message length technique, which can help the user to decide the optimal structure of the model. In this thesis we also demonstrate the applicability of the hierarchy of latent trait models in the field of document data mining.

Discriminating antigen and non-antigen using proteome dissimilarity:bacterial antigens

It has been postulated that immunogenicity results from the overall dissimilarity of pathogenic proteins versus the host proteome. We have sought to use this concept to discriminate between antigens and non-antigens of bacterial origin. Sets of 100 known antigenic and nonantigenic peptide sequences from bacteria were compared to human and mouse proteomes. Both antigenic and non-antigenic sequences lacked human or mouse homologues. Observed distributions were compared using the non-parametric Mann-Whitney test. The statistical null hypothesis was accepted, indicating that antigen and non-antigens did not differ significantly. Likewise, we were unable to determine a threshold able to separate meaningfully antigen from non-antigen. Thus, antigens cannot be predicted from pathogen genomes based solely on their dissimilarity to the human genome.

Discriminating antigen and non-antigen using proteome dissimilarity II:viral and fungal antigens

Immunogenicity arises via many synergistic mechanisms, yet the overall dissimilarity of pathogenic proteins versus the host proteome has been proposed as a key arbiter. We have previously explored this concept in relation to Bacterial antigens; here we extend our analysis to antigens of viral and fungal origin. Sets of known viral and fungal antigenic and non-antigenic protein sequences were compared to human and mouse proteomes. Both antigenic and non-antigenic sequences lacked human or mouse homologues. Observed distributions were compared using the non-parametric Mann-Whitney test. The statistical null hypothesis was accepted, indicating that antigen and non-antigens did not differ significantly. Likewise, we could not determine a threshold able meaningfully to separate non-antigen from antigen. We conclude that viral and fungal antigens cannot be predicted from pathogen genomes based solely on their dissimilarity to mammalian genomes.

Multispectral retinal image analysis:a novel non-invasive tool for retinal imaging

Purpose-To develop a non-invasive method for quantification of blood and pigment distributions across the posterior pole of the fundus from multispectral images using a computer-generated reflectance model of the fundus. Methods - A computer model was developed to simulate light interaction with the fundus at different wavelengths. The distribution of macular pigment (MP) and retinal haemoglobins in the fundus was obtained by comparing the model predictions with multispectral image data at each pixel. Fundus images were acquired from 16 healthy subjects from various ethnic backgrounds and parametric maps showing the distribution of MP and of retinal haemoglobins throughout the posterior pole were computed. Results - The relative distributions of MP and retinal haemoglobins in the subjects were successfully derived from multispectral images acquired at wavelengths 507, 525, 552, 585, 596, and 611?nm, providing certain conditions were met and eye movement between exposures was minimal. Recovery of other fundus pigments was not feasible and further development of the imaging technique and refinement of the software are necessary to understand the full potential of multispectral retinal image analysis. Conclusion - The distributions of MP and retinal haemoglobins obtained in this preliminary investigation are in good agreement with published data on normal subjects. The ongoing development of the imaging system should allow for absolute parameter values to be computed. A further study will investigate subjects with known pathologies to determine the effectiveness of the method as a screening and diagnostic tool.