Minimum variance

and CEAUL 1 TOPICS under consideration: 1. With PORT standing for Peaks Over Random Thresholds, we first refer the classical PORT estimation of a positive extreme value index (EVI), γ, and associated estimation of the Value at Risk (VaR) at a level p, the size of the loss occurred with a small probability p. These estimators are dependent on a tuning parameter q, 0 ≤ q < 1, that makes them highly flexible. • The PORT estimators of the EVI are invariant for changes in location and scale. • The PORT estimators of VaR satisfy the adequate linear property of quantiles, χ p (s + δX) = s + δ χ p (X). 2. We also refer the most simple Minimum-Variance Reduced-Bias (MVRB) EVI estimators, and associated MVRB-VaR estimators.

In this paper, we deal with semi-parametric corrected-bias estimation of a positive extreme value index (EVI). Then, the classical EVI-estimators are the Hill estimators, based on any intermediate number k of top order statistics. But these EVIestimators are not location-invariant, contrarily to the peaks over random threshold (PORT)-Hill estimators, which depend on an extra tuning parameter q. On the basis of second-order minimumvariance reduced-bias (MVRB) EVI-estimators, we shall here consider PORT-MVRB EVI-estimators, and propose the use of a heuristic algorithm, for the adaptive choice of k and q. Applications in the fields of insurance and finance will be provided. The following files were submitted by the author for peer review, but cannot be converted to PDF. You must view these files (e.g. movies) online.

Increased use of alternative fuels and low commodity prices have contributed to the recent expansion of the U.S. ethanol industry. As with any competitive industry, some level of output price risk exists in the form of volatility; yet, no actively traded ethanol futures market exists to mitigate output price risk. This study reports estimated minimum variance cross-hedge ratios between Detroit spot cash ethanol and the New York Mercantile Exchange unleaded gasoline futures for 1-, 4-, 8-, 12-, 16-, 20-, 24-, and 28-week hedge horizons. The research suggests that a one-to-one cross-hedge ratio is not appropriate for some horizons.

A relati ely simple approach to Nonlinear Prediclive (;en 'rali:;ed Minilllum Variance (NPGMV) contr 1 is intr' duced for nOl1lin,"()J" di"cr 'Ie-lime tIIulli Jllriuh!" systems, The system i repre ented by a comhination of a stable nonlinear ubsystem where no tructure i assumed and a lin ar subsy tern that may be unstable and modell d in polynomial matriA form. The multi-step predictive control cost index to b minimised in olves both wei ht d error and control signal cosling terms. he NPCMV control law involves an as umption on the ch icc of cost-function weights to ensure the existence of a stable n nlinear clos d-Ioop oper tor. A valuable feature orthe control law is that in the asymptotic case. where th plant is linear, the 1 controller r duces to a polyn mial matri.\ version of the well known CPC c ntroller. In the limiting case when the plant is nonl inear and the cost-function is ingle tep the controller becomes equal to the polynomial matrix version of the so called Nonlinear Generalised Minimum Vuriunce controller. Th contr Iler can be implemented in a form related to a nonlinear version of the Smith Predictor but unlik thi' compensator a stabilizing control Jaw can be obtained for open-loop unstable proc sses.

In this article, a robust, stable and fast calculable controller that reduces the variance to the minimum for minimal and non-minimal phase Linear Time Invariant (LTI) system is proposed. The calculation is based on an algorithm that overcomes the complexity of conventional methods. The algorithm utilizes Diophantine Equation to obtain overparameterized transfer function polynomial forms that contribute to the variance reduction. It analytically proves that increasing the order of the parameterized controller's coefficients makes the variance converge to the minimum, while maintaining the same simplicity of calculation. Simulated examples for different LTI models support our findings.

This paper proposes a digital amplitude-phase weighting array based a minimum variance multi-frequency distortionless restriction (MVMFDR) to aviod the frequency band signal distortion in digital beamformer and too short time delay line (TDL) requirement in analoge wideband TDL array.

A time-frequency representation can highlight non-stationarities in a signal. We propose to extract subsets from the Time-Frequency Representation (TFR) for classification or recognition purposes. We developed two approaches. The first one is developed for TFRs obtained from the Short Time Fourier Transform or the gliding Minimum Variance method. The extraction of compact subsets is viewed as a segmentation of the TFR, which is performed by morphological filtering and Watershed segmentation. The second approach is developed when the TFR has been obtained using parametric estimators. We consider a hybrid estimator, the ARCAP method, and use a Kalman filter trajectory tracker to extract spectral lines. The proposed methods are illustrated by examples on natural signals : dolphin whistle acoustical signals, cavitation signals and seismic signals produced by snow avalanches.

Many studies in the area of portfolio selection have done based on trade-off among various moments especially between mean and risk of sample returns. Merton (1980) argued that the instability of portfolio weights and sampling errors are due more to estimate the amount of mean. Indeed, it is difficult to estimate the expected return from time series of realized expected return. By this paper, we first answer to this question how to eliminate the wrong effect of mean sample returns instead of ignoring it from calculating portfolio weights and how would be the relation of other moments after this consideration. Second, the substantial evidence from experiments shows that hide information exposes ambiguity aversion to investor’s behavior, and hence decision making under ambiguity for portfolio choice has led to improving tractability of the main features of asset returns. By considering the volatility of utility preferences as ambiguity then individuals prefer to stabilize their utilit...

New information on the presence and relative abundances of 41 reef-building (zooxanthellate) coral species at 11 eastern Paci®c and 3 central Paci®c localities is examined in a biogeographic analysis and review of the eastern Paci®c coral reef region. The composition and origin of the coral fauna and other reef-associated taxa are assessed in the context of dispersal and vicariance hypotheses. A minimum variance cluster analysis using coral species presence±absence classi®cation data at the 14 localities revealed three eastern Paci®c reefcoral provinces: (1) equatorial ± mainland Ecuador to Costa Rica, including the Gala pagos and Cocos Islands; (2) northern ± mainland Me xico and the Revillagigedo Islands; (3) island group ± eastern Paci®c Malpelo Island and Clipperton Atoll, and central Paci®c Hawaiian, Johnston and Fanning Islands. Coral species richness is relatively high in the equatorial (17±26 species per locality) and northern (18±24 species) provinces, and low at two small oshore island localities (7±10 species). A high proportion (36.6%, 15 species) of eastern Paci®c coral species occurs at only one or two localities; of these, three disappeared following the 1982±83 ENSO event, three occur as death assemblages at several localities, and ®ve are endangered with known populations of ten or fewer colonies. Principal component analysis using ordinal relative density data for the 41 species at the 14 localities indicated three main species groupings, i.e., those with high, mid, and narrow spatial distributions. These groupings correlated with species population-dynamic characteristics. These results were compared with data for riverine discharges, ocean circulation patterns, shoreline habitat characteristics, and regional sea surface temperature data to help clarify the analyses as these measures of environmental variability aect coral community composition. Local richness was highest at localities with the highest environmental variability. Recent information regarding the strong af-®nity between eastern and central Paci®c coral faunas, abundance of teleplanic larvae in oceanic currents, high genetic similarity of numerous reef-associated species, and appearances of numerous Indo-west Paci®c species in the east Paci®c following ENSO activity, suggest the bridging of the east Paci®c ®lter bridge (formerly east Paci®c barrier). Ecology obviously plays a crucial role and, although its eects cannot be detected in compilations of entire faunas, must be taken into account if further progress is to be made in our understanding of the history of diversity and the factors that aect speciation. (G.J. Vermeij 1993) Coral Reefs (2000) 19:1±23

A structured unsegregated cybernetic model able to describe a diauxic growth phenomena of cells colony in aerobic condition. In this paper, the model has been proven in the simulation of the behavior of a batch and fed-batch bioreactors achieving satisfactory results. For simulating fed batch system, at first, the optimal substrate feeding in to bioreactor has been obtained by solving optimal control problem. Applying this profile to process model shows that ethanol concentration is much less in comparison with a constant feed rate. Also, because in many fermentation processes, oxygen transfer is the rate limiting step, so for preventing from oxygen starvation that causes ethanol production, oxygen mass transfer coefficient is simulated on the function of impeller speed and air flow rate and then, control of oxygen concentration by PI and GLC controller have been considered. Results show that two controllers have a same performance, but because of simpler structure in PI controllers, PI controller is better than GLC for this process. Also cell concentration without using controller have been brought and shown that productivity is smaller in comparison with using controllers.

We describe the on-board electronics chain and the on-ground data processing pipeline that will operate on data from the Herschel-SPIRE photometer to produce calibrated astronomical products. Data from the three photometer arrays will be conditioned and digitised by on-board electronics and sent to the ground with no further on-board data processing. On the ground, the data pipeline will process the data from point source, jiggle-map, and scan-map observations in a fully automatic manner, producing measured flux densities (for point source observations) or maps. It includes calculation of the bolometer voltages from the raw telemetry, glitch removal, and corrections for various effects including time constants associated with the detectors and electronics, electrical and optical crosstalk, detector temperature drifts, flatfielding, and non-linear response of the bolometers to strong sources. Flux density calibration will be with respect to standard astronomical sources with the plan...