Neste trabalho, considerou-se o caso da análise de covariância linear para ensaios com um fator levando-se em conta urna equação de regressão linear específica para cada tratamento. Para tanto, adotou-se o modelo estatístico (descrito na dissertação). Com base neste modelo, foram determinados: o sistema de equações normais; as estimativas dos parâmetros envolvidos no modelo; a matriz de dispersão para as estimadores dos coeficientes de regressão covariante; os componentes de variância para os efeitos considerados fixos do modelo; os critérios para os testes das hipóteses de nulidade para efeito de regressão covariante, homogeneidade dos coeficientes de regressão e efeitos ajustados dos·tratamentos; como também, os critérios para cornparações múltiplas pelo métodos de TUKEY·e teste t com base nas funç6es lineares estimáveis para efeitos dos tratamentos. / This paper refers to the linear covariance analysis for experiments with one factor when these have different linear regression equations within the treatments. Under these circunstances the following intes were determined; the normal equations system; the estimates of the envolved parameters in the model; the dispersion matrix for the estimates of the covariant regression coefficient; the variance components for the effects considered fixed in the model; the criteria for the null hypothesis test for covariants regression effects, the homogenity of the regression coefficients and the effects adjusted to the treatments; likewise the criteria for the multiple comparations

The goal of this paper is to do some basic proofs for lasso and have a deep understanding of linear regression. In this paper, firstly I give a review of methods in linear regression, and most concerns with the method of lasso. Lasso for ¡®least absolute shrinkage and selection operator¡¯ is a regularized version of method adds a constraint which uses norm less or equal to a given value t. By doing so, some predictor coefficients would be shrank and some others might be set to 0. We can attain good interpretation and prediction accuracy by using lasso method. Secondly, I provide some basic proofs for lasso, which would be very helpful in understanding lasso. Additionally, some geometric graphs are also given and one example is illustrated.

Student Number : 9306442N - PhD thesis - School of Electrical and Information Engineering - Faculty of Engineering and the Built Environment / This thesis describes the design, construction and evaluation of a prototype modified tubular linear synchronous motor. The linear motor has a long static primary and a short permanent magnet mobile secondary. The design is unique in that a tubular topology has been modified to allow access to the moving secondary section. This modification means that this design can be used in long distance applications. The application for this research is the use of linear motors in rope-less vertical trans- portation systems in ultra deep level mines. The design of the linear motor has been analysed from an electrical, magnetic, mechanical and thermal perspective. Finite Element Analysis was used to predict the performance of the linear motor. The design of the secondary section has been optimised to produce the greatest possible thrust force while reducing the effects of the cogging forces. The linear motor has been tested extensively and the results correlate with theoretical predictions from the Finite Element Analysis. This project proves that the modified tubular lin- ear synchronous motor is a viable technology that can be used in rope-less vertical transportation systems.

A novel control strategy for segmented long-stator permanent magnet linear synchronous motors is presented that is particularly suited to ropeless hoisting. The stator (primary) is segmented into modular sections that are shorter than the moving reaction plate (secondary) and a single voltage source inverter per conveyance is used to power the energised sections. Efficient partial excitation of the stator is achieved with an intermediate switching layer placed between the stator sections and the inverter. The common voltage supplied to the sections is controlled with any standard motor vector control technique. Computer simulation and a purpose-built laboratory prototype show the strategy to be effective. There is scope for future research into full sensorless control of the proposed arrangement and a method of further modulating the supply voltage for those stator sections that are only partially covered by the reaction plate, particularly when operating at higher speeds. Stemming from initial efforts to mathematically model the linear motor, a novel electromagnetic modelling approach was formulated. This hybrid magnetic-equivalentcircuit (MEC) and finite element method (FEM) approach is suited to efficient dynamic simulation and is explicated with a worked example. The method yields accurate results when compared with pure FEM. An MEC formulation is used for highly-permeable polygonal regions, whereas the magnetic field outside these regions is solved, by superposition, as the combined effect of boundary conditions that interface with those MEC regions and any magnetic sources. The required parameter sets for solving the field outside the regions represented as MECs are purely dependent on the geometry of the problem space and are thus precomputed. The force acting on a chosen group of moving components is calculated using an approximation of the Maxwell Stress Tensor method. Future research could address the present limitation that the highly-permeable regions are meshed using exclusively-rectangular flux tube elements. A dynamic simulation model that caters for the discontinuities of a segmented stator was ultimately derived using an extended space-vector approach and its implementation as a generic Simulink R blockset is detailed. The motor parameters are determined from static FEM solutions over the range of reaction plate displacement, assuming no magnetic saturation.

A dissertation submitted to the Faculty of Science, University of the Witwatersrand, Johannesburg, in fulfilment of requirements for the degree of Master of Science. September 2014. / A new method of solution to the problem of nding the minimum `1 norm solution to an algebraic underdetermined linear system is developed. The new method is a geometrically clear, primal method. Like some existing methods, the new method can be logically divided into two parts. A number of new techniques are suggested in this part of the algorithm, including an iterative ascent procedure. In the second part of the solution process, the particular solution obtained in the rst part is iteratively improved. We have developed a number of new techniques here corresponding to both single and multi-element exchange procedures. Central to the new method is the development of descent criteria for a direction vector, and the stopping condition.The performance of our algorithm is also compared with two well-known methods from the literature. Our method is shown to be much superior to these well known-methods with respect to both the number of iterations and the wall-clock time required. The iterative computational complexity of the new method also compares favourably with most well-known methods. A geometric heuristic is developed for initial active constraint set selection and a number of theoretical results are given. The heuristic stands to be much more valuable if the results presented herein can be generalised

O planejamento de uma empresa agrícola se faz sob condições de risco. Ao se utilizar a programação linear como instrumento auxiliar no processo de decisão rural, o fator risco deve ser considerado com o objetivo de melhor representar a realidade enfrentada pelo agricultor. O trabalho avalia a viabilidade de se utilizar a simulação Monte Carlo como forma de incorporação da aleatoriedade dos coeficientes nos modelos de programação linear. Os resultados mostram que o método proposto é capaz de fornecer informações não apresentadas por outros modelos e, para alguns casos, poderá se apresentar como único método adequado de análise. / Farm planning is done under risky conditions. When using linear programming as an auxiliary tool in the rural decision process, risk factor must be considered in order to better represent the reality faced by the farmer. The work evaluates the feasibility of using a Monte Carlo simulation approach to incorporate the randomness of the coefficients in the linear programming models. The results show that the proposed method is able to provide information not presented by other models and, in some cases, may become the only appropriate method of analysis.

Lau Yuk Hong. / Bibliography: leaves 63-64 / Thesis (M.Ph.)--Chinese University of Hong Kong, 1984

by To Tze-ming. / Bibliography: leaves 92-93 / Thesis (M.Ph.)--Chinese University of Hong Kong, 1985

by Lam Man Kin. / Thesis (M.Phil.)--Chinese University of Hong Kong, 1990. / Bibliography: leaves 69-71. / Chapter Chapter 1 --- Introduction --- p.1 / Chapter Chapter 2 --- Generalized Linear Models --- p.7 / Chapter §2.1 --- The Model --- p.7 / Chapter §2.2 --- Estimation of the Parameters --- p.10 / Chapter Chapter 3 --- Tests for the Dispersion Parameters --- p.15 / Chapter S3.1 --- Three Asymptotically Equivalent Tests based on MLE --- p.15 / Chapter §3.2 --- Application of the Tests for the Dispersion Parameters --- p.22 / Chapter Chapter 4 --- Finite Sample Study of the Three Tests based on Simulation --- p.29 / Chapter §4.1 --- Introduction --- p.29 / Chapter §4.2 --- The Simulation --- p.31 / Chapter §4.3 --- The Results --- p.34 / Chapter Chapter 5 --- An Example --- p.38 / Chapter Chapter 6 --- Conclusions and Discussions --- p.42 / Tables / References

by Wai-Chiu Cheung. / Thesis (M.Phil.)--Chinese University of Hong Kong, 1992. / Includes bibliographical references (leaves 117-119). / Introduction --- p.1 / Chapter Chapter 1 --- Topological Riesz Spaces --- p.5 / Chapter 1.1 --- Locally convex spaces --- p.5 / Chapter 1.2 --- Ordered vector spaces and Riesz spaces --- p.10 / Chapter 1.3 --- Locally convex Riesz spaces --- p.16 / Chapter 1.4 --- Banach lattices --- p.23 / Chapter Chapter 2 --- Operator Modules and Ideal Cones --- p.32 / Chapter 2.1 --- Operator Modules and Ideal Cones on Banach lattices --- p.32 / Chapter 2.2 --- Half-injective hull and half-surjective hull of operator modules and ideal cones --- p.38 / Chapter 2.3 --- Topologies determined by operator modules and ideal cones --- p.49 / Chapter 2.4 --- Bornologies determined by operator modules and ideal cones --- p.56 / Chapter Chapter 3 --- Banach lattices of operators --- p.63 / Chapter 3.1 --- Cone absolutely summing maps --- p.64 / Chapter 3.2 --- Compact operators on Banach lattices --- p.72 / Chapter 3.3 --- PL-compact operators and locally order precompact operators --- p.85 / Chapter 3.4 --- Almost order bounded sets and semicompact operators --- p.100 / Ref erences --- p.117