[pt] A previsão para dados de alta freqüência é fundamental para a segurança e confiabilidade da operação do sistema elétrico. Os métodos de amortecimento exponencial, em particular o método de Holt-Winters e suas variações, são apropriados para este contexto devido à sua alta adaptabilidade e robustez. Este trabalho visa à previsão a cada quinze minutos para sete dias à frente, isto é, 672 passos à frente, para dados de energia elétrica de uma concessionária da região sudeste do Brasil. Para tanto, será utilizado o método de Holt-Winters com dois ciclos, que foi proposto recentemente por J. Taylor. Além disso, será incluído o tratamento de feriados e a influência da temperatura, que serão aplicadas de forma exógena ao modelo. / [en] The forecast for high frequency data is fundamental for the safety and reliability of the electric power system operation. The methods of exponential smoothing, particularly the Holt-Winters approach and its variations, are appropriate for this type of data due to its highly adaptability and robustness. This work seeks to produce forecasts, every fifteen minutes, for a time horizon of seven days ahead, that is, 672 steps of fifteen minutes ahead, for a brazilian load series of an important distribution utility located in the southeast region of the country. It is considered the two-cycle version of the Holt-Winters approach as proposed recently by J. Taylor, including some improvements, such as, the treatment of bank holidays and the influence of temperature, acting as exogenous inputs.

Acoustic filters for signal filtering are used in wireless technologies operating at gigahertz frequencies for communication systems such as next generation cell phones. Multilayered porous silicon structures have been fabricated from silicon wafers to create the Bragg mirror section of a bulk acoustic wave filter. These porous silicon multilayers have been designed for use from 500 MHz – 20 GHz with primary focus on frequencies at 1 GHz. The porous silicon multilayers consist of alternating layers of high and low acoustic impedance layers on a bulk silicon substrate. They are fabricated using electrochemical etching where the current density during the etch determines the porosity and hence acoustic impedance of each layer. Bragg mirrors, FabryPerot filters, microcavities and rugate filters can be produced in this way due to the control of the tuneable porosity profile throughout the structure. The porosity of the layer modifies the elastic constants of the layer such as the Young’s modulus and hence the velocity of the bulk acoustic waves travelling through it. The behaviour of bulk acoustic waves through silicon is known but in order to fabricate porous silicon acoustic filters, the dependence of the longitudinal wave velocity as a function of porosity must also be known. This has been studied using acoustic transmission measurements on single porous silicon layers and then extended to multilayered structures. Rugate filters are single frequency filters that have not previously been studied for acoustic applications. In this study the first acoustic rugate filters have been fabricated using porous silicon material that exhibit only one stopband near 1 GHz. Bragg mirrors have been made with acoustic transmission measurements showing the locations of the stopbands. Porous silicon microcavities have also been fabricated along with filters that have apodisation functions. This work could form the basis of future efforts to produce and incorporate allSi multilayers into acoustic filters that are easily fabricated at a high level of quality and reliability that will serve to be efficient and cost effective.

Made available in DSpace on 2016-04-27T18:12:05Z (GMT). No. of bitstreams: 1 Leila Quadrado Lopes.pdf: 588645 bytes, checksum: 5388697821e4f79f1026dafd999a55b0 (MD5) Previous issue date: 2005-02-03 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / The present study aimed at applying high frequency audiometry in the frequency range from 9.000 to 20.000 Hz in individuals, in order to verify if ear, gender and age are variability sources. Method: There were selected 65 individuals (130 ears), being 22 male and 43 female in the age ranging from 20 to 60 years old, with no otological past, with normal tympanometric curves and hearing thresholds in pure tone conventional audiometry (250 8.000 Hz) and presence of acoustic reflexes. The audiometer used was a Grason Stadler GSI 61 with the Senheiser HAD-200 earphones, which allowed dB hearing threshold level. Results: there were no statistically significant differences between tested ears; there were statistically significant differences between age range, being worse the hearing thresholds with the increasing age, as well as there were statistically significant differences between hearing thresholds of female and male individuals, in which women has presented worse hearing thresholds only in 16.000 and 18.000 Hz. Nevertheless, in the other frequencies, there was a tendency of improving the hearing thresholds for female individuals. Conclusion: Despite of this procedure has not being used in the clinical routine, by evaluating the high frequencies thresholds we can get more information about the hearing system, increasing its comprehension and also providing more chances of a more precise prognostics about hearing lesions / O presente estudo teve por objetivo aplicar a audiometria de altas freqüências (9000 a 20000 Hz) visando verificar se lado da orelha faixa etária e sexo são fonte de variabilidade Método Foram selecionados 65 indivíduos (130 orelhas) 22 do sexo masculino e 43 do sexo feminino com idades entre os 20 e 60 anos sem passado otológico, com curvas timpanométricas e limiares de audibilidade na faixa de frequencia convencional (250 a 8000 Hz) dentro da normalidade e presença de reflexos acústicos Foi utilizado na pesquisa das altas freqüencias o audiometro Grason Stadler - GSI 61 e fones Senheiser HDA -200 que possibilitaram a verificação dos limiares auditivos em dB NA Resultados Após análise estatística os resultados demostraram que não houve diferenças estatisticamente significantes com relação ao lado da orelha testada Houve diferenças estatisticamente significantes com relação à faixa etária, sendo piores os limiares auditivos conforme o aumento da idade bem como houve diferencas estatisticamente significantes com relacao à faixa etária sendo piores os limiares auditivos conforme o aumento da idade bem como houve diferencas estatisticamente significantes nos limiares auditivos dos indivíduos dos sexos feminino e masculino sendo que mulheres apresentaram limiares auditivos piores do que homens somente nas frequencias de 16000 e 18000 Hz Entretanto nas demais freqüencias houve uma tendencia à melhora nos limiares para os sujeitos do sexo feminino Conclusao Apesar deste procedimento ainda não ser utilizado na rotina clínica atraves da avaliação das altas freqüências poderemos obter mais informacoes a respeito do sistema auditivo, aumentando assim a compreensão deste e também possibilitando maiores chances de chegarmos a um prognóstico mais preciso sobre as lesões auditivas

For financial forcasting of crisis new concepts from disciplines dissimilar to economics are looked for by financial experts. The branch of econophysics using theories of natural sciences is significant. The meaning of this work is to point out one of many methods applied to financial data with help of the theory of turbulence of fluids and deterministic chaos. We provide a parallel analysis of high frequency financial time series of a stock index and velocities of a turbulent fluid. This work concerns the use of concepts from statistical mathematics, probability theory and scaling. We find differences of both studied systems but the methodologies of natural diciplines can be also applied to financial data.

As opposed to current Intraoperative Ultrasound (IOUS) systems and their relatively large probes and limited superficial high frequency imaging, the use of a biopsy needle with an integrated transducer that is capable of minimally invasive and high-resolution ultrasound imaging is proposed. Such a design would overcome the compromise between resolution and penetration depth which is associated with the use of a probe on the skins surface. It is proposed that during interventional procedures, a transducer array positioned at the tip of a biopsy needle could provide real-time image guidance to the clinician with regards to the needle position within the tissue, and aid in the safe navigation of needles towards a particular target such as a tumour in tissues such as the breast, brain or liver, at which point decisions surrounding diagnosis or treatment via in vivo tissue characterisation could be made. With this objective, challenges exist in the manufacturing these miniature scale devices and theirincorporation into needle packages. The reliable realisation of miniature ultrasound transducer arrays on fine-scale piezoelectric composites, and establishing interconnects to these devices which also fit into suitably sized biopsy needles are two such hurdles. In this thesis, the fabrication of miniature 15 MHz ultrasound transducers is presented. The first stage of development involved the production of single element transducers in needles ~2 mm inner diameter, using various piezoelectric materials as the active material. These devices were tested andcharacterised, and the expertise developed during their fabrication was used as the foundation upon which to design a wafer-scale fabrication process for the production of multiple 15 MHz transducer arrays. This process resulted in a 16 element 15 MHz array connected to a flexible printed circuit board and integrated into a breast biopsy needle. Characterisation tests demonstrated functionality of each of the 16 elements, both individually and combined as an array. To explore potential applications for these devices, the single element transducers were tested in fresh and Thiel embalmed cadaveric brain tissue. Plasticine targets were embedded in these brain models and the needle transducers were tested as navigational real-time imaging tools to detect these targets within the brain tissue. The results demonstrated feasibility of such devices to determine the location of the target as the needle devices were advanced or withdrawn from the tissue, showing promise for future devices enabling neurosurgical guidance of interventional tools in the brain. The application of breast imaging was also considered. Firstly, Thiel embalmed cadaveric breasts were assessed as viable breast models for ultrasound imaging. Following this, anatomical features, with diagnostic significance in relation to breast cancer i.e. axillary lymph nodes and milk ducts, were imaged using a range of ultrasound frequencies (6 – 40 MHz). This was carried out to determinepotential design parameters (i.e. operational frequency) of an interventional transducer in a biopsy needle probe which would best visualise these features and aid current breast imaging and diagnosis procedures.

Magister Public Health - MPH / Ototoxic induced hearing loss is a common adverse event related to aminoglycosides used in Multi Drug Resistant -Tuberculosis treatment. Exposure to ototoxic drugs damages the structures of the inner ear. Symptomatic hearing loss presents as tinnitus, decreased hearing, a blocked sensation, difficulty understanding speech, and perception of fluctuating hearing, dizziness and hyperacusis/recruitment. The World Health Organization (1995) indicated that most cases of ototoxic hearing loss globally could be attributed to treatment with aminoglycosides. The aim of the study was to determine the proportion of DR-TB patients initiated on treatment at three decentralized sites during a defined period (1st October to 31st December 2015) who developed ototoxic induced hearing loss and the corresponding risk factors, whilst receiving audiological monitoring with an extended high frequency audiometer (KUDUwave). A retrospective cross-sectional study was conducted. Cumulatively across the three decentralized sites, 69 patient records were reviewed that met the inclusion criteria of the study. The mean age of the patients was 36.1, with a standard deviation (SD) of 10.7 years; more than half (37) were female. Ototoxicity , a threshold shift, placing patients at risk of developing a hearing loss was detected in 56.5% (n=39)of patients and not detected in 30.4%(n=21).The remaining 13,1% (n=9)is missing data. As a result, the regimen was adjusted in 36.2% of patients. . From the 53 patients who were tested for hearing loss post completion of the injectable phase of treatment, 22.6% (n=12) had normal hearing, 17.0 % (n=9) had unilateral hearing loss, and 60.4% (n=32) had bilateral hearing loss. Therefore, a total of 41 patients had a degree of hearing loss: over 30% (n=22)had mild to moderate hearing loss, and only about 15% (n=11)had severe to profound hearing loss. Analysis of risk factors showed that having ototoxicity detected and not adjusting regimen significantly increases the risk of patients developing a hearing loss. The key findings of the study have shown that a significant proportion of DR-TB patients receiving an aminoglycoside based regimen are at risk of developing ototoxic induced hearing loss, despite receiving audiological monitoring with an extended high frequency audiometer that allows for early detection of ototoxicity (threshold shift).

Roughly all the previous empirical research, focusing on the information effects on volatility, has investigated the volatility dynamics during and after the release of public information. Researchers use ARCH-type or realized volatility models and they proxy public information by market news announcements. So far, studies focusing on the effect of noise or technical trading on volatility have been limited to theoretical results without any empirical evidence. Technical trading is trading based on technical signals. As a consequence, the aim of this dissertation is to answer to the following question: how does foreign exchange volatility behave, in the very short term, around public information and technical signals ? To answer to this question we study the volatility dynamics before, during and after public news announcements and technical chart pattern signals. In order to meet this objective, we implement different methodologies specific to the different chapters of the dissertation. Each chapter tries to answer to a sub-question emerging from the main question of the thesis. This thesis contributes to the empirical finance literature on intradaily exchange rate volatility as follows. First we present evidence that volatility increases in the pre-announcement period of scheduled news. Second, we show that foreign exchange dealers quoting activity reacts to news announcements and it conveys useful information. The third contribution consists in presenting a new approach to recognize technical chart patterns from a time series, and shedding light on the predictive success of the technical chart signals. Finally, the last contribution consists in the finding that technical signals, considered by economists as "noise", have a significant effect on volatility.

The bid-ask spread costs consist of three components, which include order processing costs, inventory-holding costs, and adverse selection costs. In this paper, we model the inventory-holding costs of the bid-ask spread by option hedging. Theinventory-holding costs are hedged by call or put option positions. Since trades deal with the adverse selection traders are unobservable. We treat it as a latent variable, and Expected-Maximization (EM) algorithm are applied to estimate the related parameters of the model. Simulation studies are performed for several different models. Empirical results of NYSE high frequency data show that the proposed model are obtain appropriate parameter estimation when the returns satisfied normality assumption.

none

The Dynamic Conditional Correlation (DCC) model proposed by Engle (2002) has become one of the most popular models for the analysis of multivariate financial time series. Yet, the impact of temporal aggregation on the DCC estimates has not yet been rigorously investigated. This thesis examines the changes of DCC estimates when the intraday returns are aggregated from 5-minutes to 270-minutes returns using Taiwanese eight industry index returns from Jan. 2, 2004 to Dec. 31, 2006. Our empirical analysis finds that dynamic correlation coefficients between the 8 industry index returns are all positive and time-varying. Further, Electronic and Building indices seem to have high correlation with other industry indices whereas plastics has a lower correlation with others. What is more important, all return series have higher conditional correlation for lower frequencies. In other words, temporary aggregation will increase the conditional correlation. This thesis also seeks to categorize the loan accounts of small- and medium-scale corporations according to their respective business sectors and calculate the monthly returns and standard deviation of the bank loans according to the groups of sample of credit records from each sector, with the purpose of establishing the efficient frontier of the loan combinations of the banks and estimation the dynamic conditional correlation to discover the optimal crediting policy. It is expected that the discussion using the model presented in the thesis may provide the basis for financial institutions as they establish their respective crediting policies.