This thesis describes the development of a decision support tool for unplanned maintenance of aircraft at ramp time during airport operations. Ramp time is the time between an aircraft arrival and its next departure. Clearance of an aircraft for flight is controlled by aviation regulations. Therefore decisions regarding maintenance are taken by engineers who have to comply with the regulations that are governed outside the organizational structure of the airline. Unplanned maintenance also often disrupts the normal operational scheduling and leads to significant costs. Therefore, the decision support tool must include the relevant aviation regulations, be capable of rescheduling to minimise disruption and be able to optimise solutions based on cost. In this project an aircraft schedule is used to demonstrate the procedures. An assumed fleet of six airplanes fly between three cities. Consultation with aviation experts ensured the size of the fleet and operations are realistic. A regulation database was developed based on the Master Minimum Equipment List (MMEL) for the aircraft, and a computer programme was developed to provide different options that comply with the regulations and take into account scheduling disruption and costs. In certain cases the regulations allow an aircraft to fly with some components inoperable so long as backup systems can perform the tasks. It is possible then to postpone the maintenance until the aircraft arrives at a properly equipped airport, or until a longer scheduled stopover reduces the disruption to operations. To address the engineering aspects of the project, maintenance of a single component that appears in the MMEL for the chosen aircraft is considered. To plan maintenance following a failure, the cause of the failure needs to be identified. Only then can the resources and time required to repair the defect be defined. The programme validation has confirmed it is able to balance different aspects of decisions related to unplanned aircraft ramp maintenance. Although the programme is based on an assumed fleet operation, the structure of the programme will allow it to be applied to other fleet and route configurations.

This dissertation is based upon a research project designed to identify and evaluate alternative agricultural systems which are applicable to the arid environment of Saudi Arabia within a multiobjective context. The four systems are: traditional; conventional; aridity-oriented; and, controlled environment. These systems differ in their utilization of basic resources, the the amount and type of food they produce, the profits they generate and their compatibility with Saudi Arabian social traditions. Thus, the environmental and sociological consequences of their implementation were considered along with production and economic aspects. The procedure for evaluating the alternative agricultural systems is a computer program called ESAP (Evaluation and Sensitivity Analysis Program) which uses multi-attribute theory as an aid to decision making. Computations to determine the extent of that achievement are essentially a weighting of the variables identified as subdivisions of the goals. Decision makers are also required to give the relative values to the variables, and to select a particular utility function which describes the relationship between value and utility. The values assigned to each variable are usually presented as a range to express the users' uncertainty. Six consultants (five university professors plus the author) with varying professional backgrounds and knowledge of Saudi Arabian conditions were used as individual and collective decision makers to evaluate the four agricultural systems and their combinations. The procedure resulted in grouping these ten different alternatives (four systems plus combinations of any two) into three independent classes: I, Il and III. The grouping was based upon obtaining a clear distinction in overall score between the classes. The grouping into classes resulted in the aridity-oriented agricultural system being the only alternative in Class I. The consultants felt that this system offered the most favorable tradeoff between the economic benefits and the social and environmental factors. In general, the study indicated that the protection of natural resources and the maintenance of cultural factors should be given significant influence along with the economic factors in evaluating a particular plan of action. In utilizing such a procedure, the need for additional data and research became very evident, if there is to be better allocation of the Kingdom's agricultural resources.

Increased global competition is forcing organizations to increase their use of group decision making today. Computerized group decision support aids (CGDSAs) are being developed to improve the efficiency of these groups and to improve decision quality. Even though the use of CGDSAs has increased, very little research has been done on the evaluation of CGDSAs. The purpose of this research was to develop a model-based generalized methodology for CGDSA evaluation from the user's perspective. Two models were developed as a foundation for the CGDSA evaluation methodology. The first model was a model of group decision making and the second model was a model of computer-aided group decision making. The group decision making model was based upon a basic input-output model with the problem as the input and the selected alternative as the output. Analogous to how problems are viewed in terms of classical design of experiments, independent variables affect the outcome (problem solution the dependent variable) of the decision making process. As in design of experiments, independent variables are either noise variables or control variables. In the model presented, the independent variables are further divided into four categories (internal, external, process, and problem) in the group decision making model as a way to help develop an exhaustive list of independent variables affecting the decision making process. The generalized methodology for CGDSA evaluation mapped directly to the computer-aided group decision making model. Solution quality is measured directly or by measuring independent variables that have been previously been correlated to solution quality using standard design of experiment techniques. The generalized methodology for CGDSA evaluation was applied to the assessment of ConsensusBuilder, an example of a CGDSA. As prescribed by the CGDSA evaluation methodology, usability was also assessed and practical use considerations were followed when designing the evaluation. The value of the ConsensusBuilder evaluation for this research was that it was possible to perform a thorough evaluation of ConsensusBuilder, a CGDSA, using the CGDSA Evaluation Methodology developed in this research. In addition to the ConsensusBuilder evaluation, six different CGDSA evaluations cited in the literature were assessed in terms of the CGDSA evaluation methodology. / Graduation date: 2002

Not available / text

Over the past decade, two types of decision aids, i.e., decision support systems (DSS) and expert systems (ES), have been developed along parallel paths, showing some significant differences in their software architectures, capabilities, limitations, and other characteristics. The synergy of DSS and ES, however, has great potential for helping make possible a generalized approach to developing a decision aid that is powerful, intelligent, and friendly. This research establishes a framework for such decision aids in order to determine the elementary components and their interactions. Based on this framework, a generalized intelligent problem solving system (GIPSS) is deveolped as a decision aid generator. A relational model is designed to provide a unified logical view of each type of knowledge including factual data, modeling knowledge, and heuristic rules. In this knowledge model, a currently existing relational DBMS, with some extension, is utilized to manage each type of knowledge. For this purpose a relational resolution inference mechanism has been devised. A prototype GIPSS has been developed based on this framework. Two domain specific decision aids, COCOMO which estimates software development effort and cost, and CAPO which finds optimal process organization, have been implemented by using the GIPSS as a decision aid generator, demonstrating such features as its dynamic modeling capabilities and learning capabilities.

No description available.

This thesis is focused on a particular text-to-text generation problem, automatic summarization, where the goal is to map a large input text to a much shorter summary text. The research presented aims to both understand and tame existing machine learning models, hopefully paving the way for more reliable text-to-text generation algorithms. Somewhat against the prevailing trends, we eschew end-to-end training of an abstractive summarization model, and instead break down the text summarization problem into its constituent tasks. At a high level, we divide these tasks into two categories: content selection, or “what to say” and content realization, or “how to say it” (McKeown, 1985). Within these categories we propose models and learning algorithms for the problems of salience estimation and faithful generation. Salience estimation, that is, determining the importance of a piece of text relative to some context, falls into a problem of the former category, determining what should be selected for a summary. In particular, we experiment with a variety of popular or novel deep learning models for salience estimation in a single document summarization setting, and design several ablation experiments to gain some insight into which input signals are most important for making predictions. Understanding these signals is critical for designing reliable summarization models. We then consider a more difficult problem of estimating salience in a large document stream, and propose two alternative approaches using classical machine learning techniques from both unsupervised clustering and structured prediction. These models incorporate salience estimates into larger text extraction algorithms that also consider redundancy and previous extraction decisions. Overall, we find that when simple, position based heuristics are available, as in single document news or research summarization, deep learning models of salience often exploit them to make predictions, while ignoring the arguably more important content features of the input. In more demanding environments, like stream summarization, where heuristics are unreliable, more semantically relevant features become key to identifying salience content. In part two, content realization, we assume content selection has already been performed and focus on methods for faithful generation (i.e., ensuring that output text utterances respect the semantics of the input content). Since they can generate very fluent and natural text, deep learning- based natural language generation models are a popular approach to this problem. However, they often omit, misconstrue, or otherwise generate text that is not semantically correct given the input content. In this section, we develop a data augmentation and self-training technique to mitigate this problem. Additionally, we propose a training method for making deep learning-based natural language generation models capable of following a content plan, allowing for more control over the output utterances generated by the model. Under a stress test evaluation protocol, we demonstrate some empirical limits on several neural natural language generation models’ ability to encode and properly realize a content plan. Finally, we conclude with some remarks on future directions for abstractive summarization outside of the end-to-end deep learning paradigm. Our aim here is to suggest avenues for constructing abstractive summarization systems with transparent, controllable, and reliable behavior when it comes to text understanding, compression, and generation. Our hope is that this thesis inspires more research in this direction, and, ultimately, real tools that are broadly useful outside of the natural language processing community.

This research assisted the Yuma, Arizona citrus orchard manager in his strategic planning for achieving a low-cost position in a focused segment of the citrus industry. Citrus growers in the Yuma district are faced with major changes in their competitive environment and must adopt new strategic plans in order to continue to compete effectively in what has recently become a global industry. Since the planning horizon for new citrus orchards is in excess of 20 years, a long range planning model was developed to aid in evaluating alternative operating strategies. This research established the interrelatedness of water, nitrogen, and phosphorous relative to the yields of Valenica Oranges, Lisbon Lemons, and Redblush Grapefruit on Rough Lemon, Sour Orange, and Troyer rootstocks. A computer simulation model was used to evaluate optimal operating policies for a variety of resource prices and market conditions. The methodology utilized in development of the simulation model was unique in that it emulates individual tree performance from the time of planting until maturation. Four operating strategies were investigated and the profit maximizing and cost minimizing strategies were found to be significant. Evaluation of market selling prices indicated that the profit maximizing strategy was optimal except at very low market prices where the cost minimization strategy was optimal. Price sensitivity for water and fertilizer resources was investigated. Operating strategies were not affected by water price increases over the foreseeable future, however, price changes in nitrogen and phosphorous were found to affect the optimal operating strategy primarily through the substitution of manure in the system. Existing horticultural practices in the Yuma growing area were confirmed by the research. Additional optimal operating strategies were suggested relative to market prices. The long run policy decision making process for orchard managers was enhanced.

Decision-making in resource allocation can be a complex and daunting task. Often there exist circumstances where there is no clear optimal path to choose, and instead the decision maker must predict future need and allocate accordingly. The application of resource allocation can be seen in many organizations, from military, to high end commercial and political, and even individuals living their daily life. We define resource allocation as follows: the allocation of owner’s assets to further the particular cause of the owner. We propose two ways that computers can assist with the task of resource allocation. Firstly they can provide decision support mechanisms, with alternate strategies for the allocations that might not have been previously considered. Secondly, they can provide training mechanisms to challenge human decision makers in learning better resource allocation strategies. In this research we focus on the latter, and provide the following general hypothesis: Coevolutionary algorithms are an effective mechanism for the creation of a computer player for strategic decision-making games. To address this hypothesis, we present a system that uses coevolution to learn new strategies for the resource allocation game of TEMPO. The game of TEMPO provides a perfect test bed for this research, as it abstracts real-world military resource allocation, and was developed for training Department of Defence personnel. The environment created allows players to practice their strategic decision-making skills, providing an opportunity to analyse and improve their technique. To be truly effective in this task, the computer player the human plays against must be continuously challenging, so the human can steadily improve. In our research the computer player is represented as a fuzzy logic rule base, which allows us investigation into the strategies being created. This provides insight into the ways the coevolution addresses strategic decision-making. Importantly, TEMPO also gives us an abstraction of another component of strategic decision-making that is not directly available in other games – that of intelligence (INTEL) and counter intelligence (CI). When resource allocation is occurring in a competitive circumstance, it is often beneficial to gain insight into what your opponent is doing through intelligence. In turn, an opponent may seek to halt or skew the information being gained. The use of INTEL and CI in TEMPO allows research into the effects this has on the resource allocation process and the coevolved computer player. The development of a computer player for the game of TEMPO gives us endless possibilities of research. In this research, we have focused on the creation a computer player that can provide a fun and challenging environment for humans learning resource allocation strategies. We investigate the addition of memory to a coevolutionary algorithm for strategy creation. This includes mechanisms to select memory individuals for evaluation of coevolutionary individuals. We describe a successful strategy of selection, based on the way a human’s short and long term memory works. We then investigate the use of INTEL and CI in the game of TEMPO, and the way it is used by the coevolved computer players. Through this work, we present a new version of the TEMPO game that more realistically represents INTEL and CI. Finally, we describe a process that uses coevolution to adapt to a human player real-time, to create a tailored game-play experience. This process was tested in a user study, and showed a distinct advantage through the adaptive mechanism. Overall, we have made some important discoveries, and described some limitations that leave future research open. Ultimately, we have shown that our hypothesis is an achievable goal, with an exciting future. / Thesis (Ph.D.) - University of Adelaide, School of Computer Science, 2008

Organic farming is increasingly viewed as a plausible production system for sustainable agriculture for smallholder farmers. However, there is not enough scientific evidence and knowledge to advocate certified organic farming for African smallholder farmers who face several constraints related to production, storage and marketing. The potential for organic farming for smallholder farmers, faced by these constraints, is not clearly defined. As a result, this study set out to evaluate the production potential of organic agriculture among three smallholder farmer groups. Production questions were used to investigate and evaluate the potential for organic agriculture among three smallholder farmer groups and constituted the following subproblems: · What crops can be grown in the three study areas, based on climatic data ? · Do farmers concur that these are the most suitable potential organic crops? · How useful do the farmers find the decision making tool? · What constraints threaten commercial production of the identified crops for these farmers? Participatory methodologies that included the use of Force Field Analysis, discussions and workshops were used to identify organic production constraints related to production decisions. Farmers faced constraints related to finance, capacity enhancement, technical knowledge, fencing, irrigation, and a lack of, or inappropriately trained extension officers. As a response to identified production constraints, a decision support tool was developed. Natural resource data, including climatic and agronomic data, was used to create a specially calibrated Microsoft Excel spreadsheet interface that functions as an empirical organic production decision support tool for organic and aspirant organic smallholder farmers, by providing answers for farmer-prioritised production constraints. A list of potential crops for each of the three study areas was subjected to a series of checks against suitability for climate and disease conditions and nutrient requirements. A limited supply of manure, to meet the enormously high requirements for organic production in the poor soils of these areas, is the major constraint to exclusive organic production and renders certified organic production difficult and unsustainable. Farmers disagreed with some of the crops on the list, arguing that familiar crops were rejected by the model, but they were excited by the prospects for production of “new” crops suggested as suitable by the decision support tool, but not yet grown in the study areas. End users welcomed the model and expressed the opinion that it would be useful in decision making related to organic crop production. The study concludes that, although a number of agronomically-suitable crops can grow in the study areas, organic production is restricted by rather high manure requirements, lack of compost making skills, lack of knowledge on natural pest and disease control and poorly nourished soils, leading to poor yields. The rainy season creates a disease-supporting environment, rendering organic farming risky for rain-fed smallholder farming. Risk in certified organic farming for smallholders was further exacerbated by a hardly inconducive policy environment that low literacy levels exist amongst farmers. This study is innovative for three reasons. First, farmers were true participants and drivers of the research. Second, trans-disciplinary expert seminars were attended by experts from different disciplines who critiqued the conceptualisation, design, and implementation of the study. Third, the development of a practical decision-support tool shows innovation towards solving complex smallholder farmers decisions. If organic farming is to be promoted, commitment by government is needed in order to establish policy and legislation on organic farming to direct and govern training, information provision and marketing. Intensive training and knowledge building of organic production for smallholder farmers and extension officers is critical. There are also agroecological risks associated with organic farming for smallholder farmers. Recommendations for future research include comparison between organic agriculture and conventional agriculture, where sustainability of certified organic farming and economic viability can be conducted in the South African context. Improvement of the decision making tool will require involving information technology specialists so that the tool can be installed in community centres, extension offices and other accessible places for farmers and others. / Thesis (Ph.D.)-University of KwaZulu-Natal, Pietermaritzburg, 2008.