Model formulas are the standard approach for specifying the variables in statistical models in the S language. Although being eminently useful in an extremely wide class of applications, they have certain limitations including being confined to single responses and not providing convenient support for processing formulas with multiple parts. The latter is relevant for models with two or more sets of variable, e.g., regressors/instruments in instrumental variable regressions, two-part models such as hurdle models, or alternative-specific and individual-specific variables in choice models among many others. The R package Formula addresses these two problems by providing a new class "Formula" (inheriting from "formula") that accepts an additional formula operator | separating multiple parts and by allowing all formula operators (including the new |) on the left-hand side to support multiple responses. / Series: Research Report Series / Department of Statistics and Mathematics

This master's thesis deals with the influence of The Kalevala and the Finnish language on the English writer John Ronald Reuel Tolkien and his work. The main purpose is to provide a possibly most complex description of The Kalevala's influence on Tolkien. Firstly, the background of The Kalevala, the role of Elias Lönnrot during its birth and its importance, and secondly, Tolkien, his work and his discovery of The Kalevala are introduced. The main part of the thesis focuses on Tolkien's inspiration by The Kalevala out of three different perspectives. The first one is the influence of the Finnish language on Quenya, one of Tolkien's fictional languages. The second one is The Kalevala's importance for Tolkien from a general point of view - especially the influence of Lönnrot's role, The Kalevala's mythology, several abstract aspects of The Kalevala led by the importance of singing and The Kalevala's structure is crucial. The third one is Tolkien's inspiration by The Kalevala from a concrete point of view. Particularly the importance of The Kalevala's character Väinämöinen and the story of the Sampo is significant here.

Ingeniero Civil Eléctrico / En el presente trabajo de título se desarrolla un sistema de detección de objetos (localización y clasificación), basado en redes neuronales convolucionales (CNN por su sigla en inglés) y dos métodos clásicos de machine learning como Random Forest (RF) y Support Vector Machines (SVMs). La idea es mejorar, con los mencionados clasificadores, el rendimiento del sistema de detección conocido como Faster R-CNN (su significado en inglés es: Regions with CNN features). El sistema Faster R-CNN, se fundamenta en el concepto de region proposal para generar muestras candidatas a ser objetos y posteriormente producir dos salidas: una con la regresión que caracteriza la localización de los objetos y otra con los puntajes de confianza asociados a los bounding boxes predichos. Ambas salidas son generadas por capas completamente conectadas. En este trabajo se interviene la salida que genera los puntajes de confianza, tal que, en este punto se conecta un clasificador (RF o SVM), para generar con estos los puntajes de salida del sistema. De esta forma se busca mejorar el rendimiento del sistema Faster R-CNN. El entrenamiento de los clasificadores se realiza con los vectores de características extraídos, desde una de las capas completamente conectadas del sistema Faster R-CNN, específicamente se prueban las tres que contempla la arquitectura, para evaluar cuál de estas permite obtener los mejores resultados. Para definir, entre otras cosas, el número de capas convolucionales a utilizar y el tamaño de los filtros presentes en las primeras capas del sistema Faster R-CNN, se emplean los modelos de redes convolucionales ZF y VGG16, estas redes son solamente de clasificación, y son las mismas ocupados originalmente. Para desarrollar los sistemas propuestos se utilizan distintas implementaciones o librerías para las cuales se dispone de su código de forma abierta. Para el detector Faster R-CNN se utiliza una implementación desarrollado en Python, para RF se comparan dos librerías: randomForest escrita en R y scikit-learn en Python. Por su parte para SVM se utiliza la librería conocida como LIBSVM escrita en C. Las principales tareas de programación consisten en desarrollar los algoritmos de etiquetado de los vectores de características extraídos desde las capas completamente conectadas; unir los clasificadores con el sistema base, para el análisis \textit{online} de las imágenes en la etapa de prueba; programar un algoritmo para el entrenamiento eficiente en tiempo y en memoria para SVM (algoritmo conocido como hard negative mining) Al evaluar los sistemas desarrollados se concluye que los mejores resultados se obtienen con la red VGG16, específicamente para el caso en que se implementa el sistema Faster R-CNN+SVM con kernel RBF (radial basis function), logrando un mean Average Precision (mAP) de 68.9%. El segundo mejor resultado se alcanza con Faster R-CNN+RF con 180 árboles y es de 67.8%. Con el sistema original Faster R-CNN se consigue un mAP de 69.3%.

In "On Fairy Stories" J.R.R. Tolkien writes that joy is the "mark of the true fairy- story." Tolkien believed that joy was the defining characteristic of the genre. This joy is not just apparent in the happy ending of the fairy tale, but also in the manner in which the plot and characters show theories of joy, and the way the text itself creates joy in the reader. This paper will explore Tolkien's creation of brightness, hope, and wonder, and how these instances express a theory of joy. First I will look at the different types of joy in Tolkien's work, then the more general theories that these types express, and finally the effect the joy in the story has on the reader. / by Natasha Minnerly. / Thesis (M.A.)--Florida Atlantic University, 2013. / Includes bibliography. / Mode of access: World Wide Web. / System requirements: Adobe Reader.

Thesis (M.B.A.)--Boston University / PLEASE NOTE: Boston University Libraries did not receive an Authorization To Manage form for this thesis or dissertation. It is therefore not openly accessible, though it may be available by request. If you are the author or principal advisor of this work and would like to request open access for it, please contact us at open-help@bu.edu. Thank you. / 2031-01-01

This thesis set out to describe how electronic portfolios support the social construction of knowledge by healthcare students. It will do this by drawing on the various forms and functions of the students’ electronic portfolios', their associated socially constructive processes and artefacts, and healthcare students’ views on how electronic portfolios are used on their courses. These factors are also considered alongside the impact of the sociohistorical and socio technical developments of electronic portfolios and healthcare course requirements. A broad grounded theory approach was used to generate substantive theory grounded in data gathered from a sample of UK and North American students (n=82). Data was gathered through an analysis of electronic portfolios (n=25), an online questionnaire (n=52), and in-depth interviews with students (n=16). The research found a limited number of socially constructive processes and artefacts that are used within healthcare courses. These processes and artefacts supported the social construction of knowledge within the students’ electronic portfolios. Even though healthcare students were not at ease with the limitations of their electronic portfolios they found that the pedagogical use of the electronic portfolio defined the range and extent of the learning they were required to demonstrate. A range of factors, internal and external to the electronic portfolio, determined the use of the electronic portfolios. This included the student’s confidence with using technology, the requirements of the professional bodies and the healthcare courses that incorporated them, and the skills of academics in using the electronic portfolios. The students found the electronic portfolios format allowed them to use a wider range of digital artefacts (i.e. images, video and audio) that were not available in paper portfolios. They also thought the integration of the Internet into their electronic portfolios meant that their portfolio was safe in a digital cloud based repository. The research found that despite the widespread use of Web 2.0 by most of the students in their personal lives, this was not incorporated into students’ electronic portfolios. This appears to be because of the lack of Web 2.0 functionality in the electronic portfolio software, and the lack of Web 2.0 knowledge in those implementing the electronic portfolios in the healthcare courses. This impact of limited processes and artefacts, and failure to use the potential of the Internet and Web 2.0 has a negative impact on the students’ abilities to socially construct their knowledge within electronic portfolios. Recommendations are made that future research identifies additional developments in software and hardware that can increase the socially constructive processes and artefacts that are incorporated into electronic portfolios. These developments must be done by consulting student users, software developers, educational technicians and academics. Finally, recommendations are made that the theory generated in the research is applied to larger samples across a wider range of healthcare student professions. The ongoing research will ensure that the theory continues to respond to ongoing hardware and software developments within the socio-historical and socio-technical student environment.

CD40 is a prominent member of the TNFR family due to its ability to be expressed by and regulate the fate of not only immunocytes, but also non-lymphoid cells. Previous studies have demonstrated that CD40 ligation by cell-surface presented agonists, and in particular membrane CD40L (mCD40L), caused extensive apoptosis specifically in a variety of malignant epithelial cells (including bladder and colorectal). By contrast, soluble CD40 agonists are weakly pro-apoptotic and only become significantly pro-apoptotic by pharmacological intervention. Recent work from our laboratory has shed light onto the tumour-specificity of CD40 as well as the differences in soluble versus membrane-presented agonists in terms of pro-apoptotic capacity. As the role of CD40 in prostate cancer remains unknown, the main aim of this study was to investigate the hypothesis that the CD40/CD40L dyad regulates prostate carcinoma (PCa) cell fate and to explore the mechanisms of this in a panel of well-characterised human PCa lines. In order to achieve CD40 ligation by mCD40L, a co-culture in vitro model was used, whereby target PCa cells were co-cultured with third-party (murine fibroblasts 3T3CD40L (engineered to express mCD40L). This mode of ligation was compared to agonistic CD40 antibody. Flow cytometry allowed detection of CD40 expression in a panel of PCa lines, comprising DU145, LNCaP and PC-3 cells. Apoptosis was detected using several assays, focusing on classical hallmarks of apoptosis (loss of cell membrane integrity, caspase activation, and DNA fragmentation). ELISA assays were employed for detection of pro-inflammatory cytokine secretion and spectrophotometry and flow cytometry were used for detection of ROS. Immunoblotting techniques were also standardised and utilised for the accurate and sensitive detection of intracellular proteins involved in CD40 signalling. Experiments using retroviruses were also employed to engineer CD40 expression in negative PCa cells. Ligation of CD40 caused apoptosis in DU145 cells and LNCaP cells. By contrast, CD40-ve cells PC-3 were refractory to CD40 ligation. Restoration of CD40 expression restores susceptibility to CD40 apoptosis. Importantly, receptor ligation by mCD40L, and not soluble agonist, could cause cell death, as soluble agonist (cross-linked G28-5 mAb) was not pro-apoptotic. mCD40L, but not G28-5, induced rapid secretion of pro-inflammatory cytokines IL-6, IL-8 and GM-CSF, thus CD40 killing was pro-inflammatory. CD40 induced apoptosis as evident by membrane integrity loss and DNA fragmentation, both hallmarks of apoptotic death. Yet, it was found that CD40 triggers a death type that is caspase-independent. The work showed that CD40 in PCa cells triggers death that does not involve cross-talk with the extrinsic pathway, but via a direct signal that involved the mitochondrial pathway as indicated by the induction of Bak and Bax proteins. mCD40L triggered rapid induction of TRAF1 and TRAF3 whilst TRAF2 expression was downregulated. ASK1 was activated which was subsequently followed by MKK7 but not MKK4 activation and this was followed by JNK phosphorylation. Functional inhibition experiments showed that both JNK/AP-1 and p38 are important for death induction. ROS production could not be detected upon CD40 activation and functional inhibition experiments showed ROS is not critical for CD40 mediated death in PCa cells, observations raising the possibility of ROS-independent ASK1 activation. Finally, preliminary experiments using prostate cancer stem cells (CSC), well-established ‘drivers’ of PCa, showed that CSCs were CD40+ve, however, within the time constraints of this project, it was not possible to assess whether CD40 ligation could induce CSC-targeted cell death. These findings have not only generated novel observations in terms of the ability of CD40 to induce PCa cell death, but have also added to our knowledge of the intriguingly multifaceted effects of CD40 in carcinoma cells. These fascinating observations imply that CD40, whilst engaging signalling pathways with some common intracellular mediators, its precise death pathways can differ both in their exact nature and their exact features. Moreover, in addition to providing biological evidence for the mechanisms of CD40 apoptosis, these observations may represent a promising targeted approach for PCa therapy as the ability to lead to extensive apoptosis in PCa cells. Equally importantly, by efficiently killing PCa cells and causing rapid pro-inflammatory cytokine secretion, whilst at the same time targeting what is potentially the cellular driver of carcinogenesis (CSCs), CD40-mediated killing represents a very promising potential therapeutic tool for PCa therapy in the near future.

余子恆. / 複印本. / Thesis (M.A.)--香港中文大學, 1988. / Fu yin ben. / Includes bibliographical references: leaves 176-190. / Yu Ziheng. / Thesis (M.A.)--Xianggang Zhong wen da xue, 1988.

Thesis (M.A.)--香港中文大學. / Ms. / Includes bibliographical references (leaves 191-195). / Thesis (M.A.)--Xianggang Zhong wen da xue. / 序言 --- p.1-6 / 導言 --- p.7-12 / Chapter 第一章 --- 分界問題及歸納法問題之歷史淵源 --- p.13-42 / 前言 --- p.13-14 / Chapter （一） --- 康德三分界問題 --- p.15-27 / Chapter （二） --- 休謨之歸納法問題　 --- p.28-42 / Chapter 第二章 --- 分界問題及歸納法問題之解決 --- p.43-94 / 前言 --- p.43 / Chapter （一） --- 哲學是什麼及哲學家的責任　 --- p.44-47 / Chapter （二） --- 知識論的概念 --- p.48-53 / Chapter (i) --- 知識論是什麼 --- p.48-50 / Chapter (ii) --- 心理主義之消除 --- p.50-53 / Chapter （三） --- 分界問題 --- p.54-78 / Chapter (i) --- 波柏爾的分界標準 --- p.54-64 / Chapter (ii) --- 分界標準所引起的困難 --- p.64-68 / Chapter (iii) --- 邏輯實証論的分界標準 --- p.68-71 / Chapter (iv) --- 波柏爾對於邏輯實徵論的分界標準之批判 --- p.71-78 / Chapter （A） --- 可否証性與可檢証性 --- p.71-75 / Chapter (B) --- 意義問題 --- p.75-78 / Chapter （四） --- 归納法問題　 --- p.79-94 / Chapter 第三章 --- 客观知識論之原素　 --- p.95-156 / 前言 --- p.95 / Chapter （一） --- 形而上實在論 --- p.96-104 / Chapter （二） --- 科學的目的 --- p.105-115 / Chapter （三） --- 真理論 --- p.116-126 / Chapter （四） --- 三個世界´ؤ´ؤ沒有認知主體的知識論 --- p.127-147 / Chapter （五） --- 批判主觀知識論 --- p.149-156 / Chapter 第四章 --- 關於客觀知識論的討論 --- p.157-190 / 前言 --- p.157 / Chapter （一） --- 關於知識概念的討論　 --- p.160-170 / Chapter （二） --- 關於分界標準可否証性的討論 --- p.171-180 / Chapter （三） --- 關於歸納法問題的討論 --- p.181-183 / Chapter （四） --- 關於三個世界問題的討論 --- p.184-190 / 参攷書籍 --- p.191-195

Management of fungal biofilms represents a significant challenge to oral healthcare. As a preventive approach, minimising adhesion between intra-oral devices and microorganisms would be an important step forward. Denture stomatitis (DS) is a multifactorial denture-associated inflammation of the oral mucosa where candidal biofilms are one of the contributing factors. Therefore, understanding candidal biofilms on dentures and finding novel strategies to control these biofilms are of significance. Interference with the adhesion step of biofilm formation is hypothetically effective strategy to control biofilms. To understand the relationship between denture candidal load, denture material type and C. albicans biofilm forming heterogeneity in DS, quantitative polymerase chain reaction (qPCR) molecular method and crystal violet (CV) assay were used. This study investigated two novel strategies to control C. albicans biofilms through interfering with adhesion: natural polyphenol curcumin (CUR) and modifying the topography of the denture material surface. Based on the optimised effective CUR concentrations, CUR adsorption to PMMA denture material was spectrophotometrically analysed. Based on these data, the effect of adsorbed CUR to PMMA and CUR pre-exposure on adhesion of C. albicans were assessed. The effect of CUR on Candida-Candida adhesion was investigated and the expression profile of selected adhesion and aggregation-associated genes was assessed using qPCR method. Micro/nano-fabricated polycarbonate and PMMA materials were replicated using injection and compression moulding techniques, respectively and were characterised using scanning electron microscopy (SEM). Adhesion of C. albicans on the micro and nano-scaled patterns was assessed using microscopic and qPCR molecular methods, respectively. The physical characteristics of the materials were assessed using theta tensiometer and a white light profiler. The data demonstrated that although C. albicans was detected in greater quantities in diseased individuals, it was not associated with increased biofilm biomass. Denture substrata were shown to influence biofilm biomass, with poly(methyl methacrylate) providing the most suitable environment for C. albicans to reside. Subsequent studies showed that CUR concentrations of 50 μg/ml could prevent adhesion to PMMA. This effect was enhanced by the CUR pre-treatment of yeast cells (>90% inhibition, p < 0.001). Investigation of the biological impact of CUR showed that it preferentially affected immature morphological forms (yeast and germlings), and actively promoted aggregation of the cells. Transcriptional analyses showed that CUR temporally modulated adhesion and aggregation associated genes. Finally, PMMA denture material was replicated to show nano features. These topographies influenced adhesion of C. albicans, depending on the candidal morphological form and the shape. Nano-pit spatial arrangements variably affect the adhesion of C. albicans, where SQ arrangement demonstrated a significant anti-adhesive capacity. Differential adhesin expression was observed on these surfaces, which were affected by the wettability and roughness of surfaces tested. In summary, C. albicans is an important determinant of denture disease, so preventing its adhesion and biofilm formation were worthwhile objectives. This thesis has shown that CUR molecules and SQ nano-pit topographies reduced C. albicans adhesion, demonstrating that chemical and physical inhibition strategies are useful. The data presented in this thesis showed the high potential of the novel strategies to be used against C. albicans biofilms, and encourages the further investigation of these approaches against polymicrobial denture biofilms.