Continuous change and development are essential aspects of evolving environments and applications, including, but not limited to, smart cities, military, medicine, nuclear reactors, self-driving cars, aviation, and aerospace. That is, the fundamental characteristics of such environments may evolve, and so cause dangerous consequences, e.g., putting people lives at stake, if no reaction is adopted. Therefore, learning systems need to apply intelligent algorithms to monitor evolvement in their environments and update themselves effectively. Further, we may experience fluctuations regarding the performance of learning algorithms due to the nature of incoming data as it continuously evolves. That is, the current efficient learning approach may become deprecated after a change in data or environment. Hence, the question 'how to have an efficient learning algorithm over time against evolving data?' has to be addressed. In this thesis, we have made two contributions to settle the challenges described above. In the machine learning literature, the phenomenon of (distributional) change in data is known as concept drift. Concept drift may shift decision boundaries, and cause a decline in accuracy. Learning algorithms, indeed, have to detect concept drift in evolving data streams and replace their predictive models accordingly. To address this challenge, adaptive learners have been devised which may utilize drift detection methods to locate the drift points in dynamic and changing data streams. A drift detection method able to discover the drift points quickly, with the lowest false positive and false negative rates, is preferred. False positive refers to incorrectly alarming for concept drift, and false negative refers to not alarming for concept drift. In this thesis, we introduce three algorithms, called as the Fast Hoeffding Drift Detection Method (FHDDM), the Stacking Fast Hoeffding Drift Detection Method (FHDDMS), and the McDiarmid Drift Detection Methods (MDDMs), for detecting drift points with the minimum delay, false positive, and false negative rates. FHDDM is a sliding window-based algorithm and applies Hoeffding’s inequality (Hoeffding, 1963) to detect concept drift. FHDDM slides its window over the prediction results, which are either 1 (for a correct prediction) or 0 (for a wrong prediction). Meanwhile, it compares the mean of elements inside the window with the maximum mean observed so far; subsequently, a significant difference between the two means, upper-bounded by the Hoeffding inequality, indicates the occurrence of concept drift. The FHDDMS extends the FHDDM algorithm by sliding multiple windows over its entries for a better drift detection regarding the detection delay and false negative rate. In contrast to FHDDM/S, the MDDM variants assign weights to their entries, i.e., higher weights are associated with the most recent entries in the sliding window, for faster detection of concept drift. The rationale is that recent examples reflect the ongoing situation adequately. Then, by putting higher weights on the latest entries, we may detect concept drift quickly. An MDDM algorithm bounds the difference between the weighted mean of elements in the sliding window and the maximum weighted mean seen so far, using McDiarmid’s inequality (McDiarmid, 1989). Eventually, it alarms for concept drift once a significant difference is experienced. We experimentally show that FHDDM/S and MDDMs outperform the state-of-the-art by representing promising results in terms of the adaptation and classification measures. Due to the evolving nature of data streams, the performance of an adaptive learner, which is defined by the classification, adaptation, and resource consumption measures, may fluctuate over time. In fact, a learning algorithm, in the form of a (classifier, detector) pair, may present a significant performance before a concept drift point, but not after. We define this problem by the question 'how can we ensure that an efficient classifier-detector pair is present at any time in an evolving environment?' To answer this, we have developed the Tornado framework which runs various kinds of learning algorithms simultaneously against evolving data streams. Each algorithm incrementally and independently trains a predictive model and updates the statistics of its drift detector. Meanwhile, our framework monitors the (classifier, detector) pairs, and recommends the efficient one, concerning the classification, adaptation, and resource consumption performance, to the user. We further define the holistic CAR measure that integrates the classification, adaptation, and resource consumption measures for evaluating the performance of adaptive learning algorithms. Our experiments confirm that the most efficient algorithm may differ over time because of the developing and evolving nature of data streams.

Resumo: O presente trabalho teve como objetivo desenvolver e comparar métodos para estimar perdas em aplicações de herbicidas em pré-emergência, utilizando como traçadores o herbicida tebuthiuron e o corante Azul Brilhante, e diferentes alvos artificiais para coleta de deriva e deposição. Para tanto, foram realizadas duas aplicações em pré-emergência, simultâneas, com pulverizadores de barra equipados com pontas de pulverização modelo SF 11002 espaçadas a 0,5 m e a 0,5 m de altura em relação ao solo. Uma aplicação foi realizada com o traçador herbicida tebuthiuron na formulação suspensão concentrada (Combine 500 SC) e a outra com o traçador Azul Brilhante (FD&C-1), nas concentrações de 0,73 e 0,60% (p.v-1) e com volumes de calda de 167,75 e 163,75 L.ha-1, respectivamente. Para estimar as perdas utilizaram-se bandejas de isopor contendo 100 g de solo espalhados em sua superfície (coletores de deposição com área útil de 0,044288 m2), em 100 repetições, distribuídas dentro da área de aplicação e; fios de náilon de 2 mm de diâmetro (coletores de deriva), fixados na posição vertical de 0 a 5 m de altura em hastes de ferro de 1,59 cm de diâmetro e 6 m de comprimento que foram colocados fora da área de aplicação, nos 4 lados adjacentes (nas posições NO, SO, SE e NE), enfileirados e fixados à distâncias de 1, 2, 4, 8 e 16 m da área aplicada. As alturas dos coletores de deriva foram de 0 a 1, 1 a 2, 2 a 3, 3 a 4 e 4 a 5 m, totalizando 4 repetições para cada distância, em cada área adjacente. Após a aplicação e secagem da calda aplicada, o solo e os fios de náilon retirados dos coletores foram guardados em sacos de polietileno. Para a extração dos traçadores contido no solo oriundo dos coletores de deposição, foram utilizadas 2 porções de 10 g de cada repetição, sendo uma lavada com 50 mL de água destilada, para extrair o Azul Brilhante... (Resumo completo, clicar acesso eletrônico abaixo). / Abstract: The aim of this research was to develop and compare methods to estimate herbicide spraying losses in pre emergence application. Two compounds were used as tracer (brilliant blue FD&C-1 and the tebuthiuron herbicide), and two different artificial targets to collect the drift and deposition. Two pre emergence simultaneous sprays were released from two sprayer booms with nozzles model SF 11002 (Jacto S.A.) distant 0.5 m and 0.5 m height. One of the sprayer applied tebuthiuron herbicide and the other brilliant blue. The concentrations used were 0.73% and 0.60% (w/v) the volumes rates were 167.75 L.ha-1 for the tebuthiuron sprayer and 163.75 L.ha-1 for the brilliant blue. It was used to estimate the drift loss deposition colectors made of foam (trays with 100 g of soil spread on its superfice with 0.044288 m2). A hundred trays were distributed in the application area resulting in 100 repetitions. Drift collectors (nylon threads of 2 mm diameter vertically fixed between 0-5 m height on iron stens with 1.59 cm diameter and 6 m length) were placed around the application area. The drift collectors were placed in all directions (NW, SW, SE and NE) at 1, 2, 4, 8 and 16 m distant and 0-1, 1-2, 2-3, 3-4 and 4-5 m height com 4 repetitions for each distance. After the sprays the soil and the nylon threads were put in a plastic bag. The analysis of the tracers in the soil samples were realized with 50 mL of destilated water to extract the brilliant blue and 50 mL of methyl alcohol to extract the tebuthiuron. The nylon threads were washed with 40 mL destilated water for extract both tracers. The wash solutions were analysed with spectrophotometer to quantify the brilliant blue and high-performance liquid chromatography (HPLC) to quantify the herbicide tebuthiuron. The drift detected values were analyzed in fatorial design 4x5x5. Gompertz model provided the highest determinations coefficients...(Complete abstract, click electronic address below). / Orientador: Carlos Gilberto Raetano / Coorientador: Edivaldo Domingues Velini / Mestre

Essa tese de doutorado discute em ato o conceito de experiência através de um processo de trabalho inventivo denominado « Deriva Parada ». Com metodologia inspirada em elementos dos movimentos artísticos Surrealismo e o Situacionismo, buscou tensionar a palavra cidade, operando nela uma crise de sentido. A deriva e a deambulação foram dois procedimentos artísticos que agiram na suspensão dos usuais sentidos atribuídos ao termo e dotaram-lhe de potência criativa-criadora de novos sentidos, ou até mesmo, da possibilidade transmitir o vazio de sentido. Seguindo a leitura de autores como W. Benjamin, G. Agamben, G. Didi-Hubeman, G. Bataille em torno da questão da produção da experiência após a modernidade histórica, buscou-se um percurso em trabalhos de artistas, pensadores, além da experência-tema dessa tese: fagulhas da destruição da experiência e outras formas de existir, contar e produzir memória na cidade contemporânea. / This thesis discusses in act the concept of experience trough a process of inventive work denominated « Deriva Parada ». It sought for tensioning the word city, operating over it a crisis of meaning with a methodology inspired in elements of the artistic movements Surrealism and Situationism. The drift and the strolling were two artistic procedures that acted on the suspension of the usual meanings related to this concept and gave it a creativecreator power of new meanings, or even, possibilities of transmitting the absence of meaning. Following the readings of authors as W. Benjamin, G. Agamben, G. Didi-Hubeman, G. Bataille around the issue of the experience production after historic modernity, we sought a trajectory in works of artists, thinkers and the theme-experience of this thesis: destruction sparks of the experience and others forms of existing, telling and producing memory in the contemporary city.

Testing of summing electronics and VDC A/D Cards was performed to assure proper functioning and operation within defined parameters. In both the summing modules and the VDC A/D cards, testing for minimum threshold voltage for each channel and crosstalk between neighboring channels was performed. Additionally, the modules were installed in Hall A with input signals from shower detectors arranged to establish a trigger by summing signals together with the use of tested modules. Testing involved utilizing a pulser to mimic PMT signals, a discriminator, an attenuator, a scaler, a level translator, an oscilloscope, a high voltage power supply, and a special apparatus used to power and send signal to the A/D cards. After testing, modules were obtained that meet necessary criteria for use in the APEX experiment, and the A/D cards obtained were determined to have adequate specifications for their utilization, with specific results included in the appendix.

Concept drift detection, the identfication of changes in data distributions in streams, is critical to understanding the mechanics of data generating processes and ensuring that data models remain representative through time [2]. Many change detection methods utilize statistical techniques that take numerical data as input. However, many applications produce data streams containing categorical attributes. In this context, numerical statistical methods are unavailable, and different approaches are required. Common solutions use error monitoring, assuming that fluctuations in the error measures of a learning system correspond to concept drift [4]. There has been very little research, though, on context-based concept drift detection in categorical streams. This approach observes changes in the actual data distribution and is less popular due to the challenges associated with categorical data analysis. However, context-based change detection is arguably more informative as it is data-driven, and more widely applicable in that it can function in an unsupervised setting [4]. This study offers a contribution to this gap in the research by proposing a novel context-based change detection and adaptation algorithm for categorical data, namely Fine-Grained Change Detection in Categorical Data Streams (FG-CDCStream). This unsupervised method exploits elements of ensemble learning, a technique whereby decisions are made according to the majority vote of a set of models representing different random subspaces of the data [5]. These ideas are applied to a set of concept drift detector objects and merged with concepts from a recent, state-of-the-art, context-based change detection algorithm, the so-called Change Detection in Categorical Data Streams (CDCStream) [4]. FG-CDCStream is proposed as an extension of the batch-based CDCStream, providing instance-by-instance analysis and improving its change detection capabilities especially in data streams containing abrupt changes or a combination of abrupt and gradual changes. FG-CDCStream also enhances the adaptation strategy of CDCStream producing more representative post-change models.

In parallel passages from Old and Middle English and in noun cognates from Modern English, Old English, and Modern German, the most discernible elements of language drift are gender loss and word meaning change, respectively. They can be observed, discussed, and calculated to show a definite progression toward the development of Modern English.

The effects of sublethal drift rates and carryover of glyphosate and dicamba into the next generation of seed potato cultivars Atlantic and Dakota Pearl are unknown. The objective of this research is to determine the impact of sublethal glyphosate and dicamba rates on mother and daughter chipping potato plants. Field studies were conducted in 2018 and 2019 in Oakes, ND. Herbicides were sprayed at the tuber initiation stage and consisted of dicamba (0, 20, and 99 g ae ha-1) and glyphosate (0, 40, and 197% g ae ha-1). During the year of application (2018), the combination of glyphosate at 197 g ha-1 and dicamba at 99 g ha-1 resulted in a 40% yield reduction compared to the non-treated in both cultivars. In 2019, the daughter tubers from mother tubers that were treated with glyphosate (23%) experienced a 16% reduction in marketable yield in both cultivars.

A brief review of the subject of palaeomagnetism as it affects the study of the behaviour of the earth's magnetic field and the problems of Continental Drift and Polar wander is presented, giving the reasons why a systematic palaeomagnetic study of the Cape and Karroo Systems of South Africa would be of outstanding significance. This task was vigorously tackled by sampling the Karroo System at vertical intervals of approximately 50 ft. in two separate areas, using the techniques that were then available. The results, although negative, provide material for a discussion of the possible reasons for the scattered directions of magnetization of the samples. A palaeomagnetic study of the Karroo dolerites was undertaken in an attempt to (i) determine 1 the position of the geomagnetic pole at the time of the intrusions, and (ii) possibly assess the importance of the remagnetization of Karroo sediments by the thermal effects of the younger intrusions. Samples from surface exposures in the eastern half of South Africa, from the shafts of a gold mine and from a railway tunnel were collected and studied, giving a reliable mean direction of magnetization of the Karroo dolerites of Declination= 341°, Inclination= -60°. At the commencement of the Jurassic period the geomagnetic pole relative to Southern Africa had the present day co-ordinates of Longitude 74½ 0 E, Latitude 70° S. Both normally and reversely magnetized dolerites were found and evidence in favour of a true reversal of the earthts magnetic field is advanced. It is also suggested that in the area studied, the dolerite was intruded in two distinct phases. Because the direction of magnetization of the Karroo dolerites is very close to that of the present magnetic field, it is difficult to separate samples remagnetized at the time of the intrusions from those remagnetized in the present field. However, some light is thrown on the problem of the scattered directions of magnetization of the Karroo sediments by the fact that dolerite samples collected from surface exposures are much less consistently magnetized than those from underground workings.

Within the third sector, scarce resources make it difficult for social enterprises to survive financially, become more innovative and entrepreneurial, and generally grow and develop to scale. Although there is no agreement on the definition, the extant literature tends to frame social enterprise as critical for addressing challenging social problems. Social enterprise involves some degree of profit-making, while maintaining a strong focus on social mission. The latest literature points to social enterprise as an example of a hybrid organisational form that has the potential to act as a solution to complex social problems. However, the literature also highlights the considerable tensions inherent in hybridity, as well as the substantive mission-drift that inevitably occurs. The current body of knowledge does not describe how social enterprise growth and development occurs, nor does it adequately illustrate how social enterprise hybrids can continue to meet their important social missions while generating sufficient operating income to sustain themselves. Additionally, there is a dominant view that non-profit social enterprises may face difficulties in surviving as hybrids due to the risk of mission-drift. An in-depth study of non-profit social enterprises that had already reached scale was undertaken to address this. By doing so, a contribution was made to the development of the social entrepreneurship and social enterprise fields. The theory on social enterprise ideal-types was deepened. In addition, this study provides a framework for social enterprise growth and development within this organising context. Strategy-as-practice (SaP) core concepts as influencers within an entrepreneurship-as-practice (EaP) theoretical framework, contained by a social practice theory theoretical lens. The result was a focus on daily business management and strategic development practices. Additionally, contextualisation – as the link between the micro-context in each organisation and macro environment – was studied as the context for this growth and development. / Thesis (DPhil)--University of Pretoria, 2020. / Gordon Institute of Business Science (GIBS) / DPhil / Unrestricted

No description available.