Content: We are a group company of a major department store in Japan. We are in charge of quality control and customer service of items sold by our stores. We constantly monitor our customers’ concerns, and we have been keeping customer correspondence records since 1996. We hold over 100,000 cases for all items of food, clothing, living and services. The record includes not only customer requests but also test results of each item before sales as well as reproduction test results based on customer requests. We use the database to build our quality standards for everything we sell and for the education of our sales staff. All personal information in these correspondence records are, of course, kept completely confidential. However, improving the quality of the items we sell is not simply a problem of finding solutions within our supply chain. Using compiled results, we aim to clarify the type and level of problems users experience with leather goods, leading to overall quality improvement of supply items of scale. However, we cannot perform this task alone. As a survey method, it is based on compiling all 7,000 records of leather goods over the past decade sorted by our original ‘product-code’ and ‘consultation-code’. By using compiled results, we are able to clarify what kinds of items and what kind of complaints is more often in major sales items. Additionally, by using text extraction from the content of customers’ requests and compiling them manually, totals can be determined for each use period and product color. For example, we can compare customer complaint data with test results for colour fastness. Referring to required performance items specified by ISO standards, the gap between customer requests and quality standards can be clarified. With the internationalization of purchasing networks of sales items, quality control based on setting various safety and quality standards is becoming increasingly important. In terms of the characteristics of leather, we think that there is probably a level that is impossible to realize, however, not only the problems of regulations and safety but the nature of the customer's viewpoint must also be discussed. Take-Away: We are going to present current data showing proof of what consumers actually want. Using this data, leather goods manufacturers can efficiently make products that consumers want and avoid making products that consumers don’t want. We suggest that quality standards , including ISO, should consider not only safety and regulations of environment, but consumer`s opinions as well.

Threat modeling for information technology infrastructure can be done using shortest path algorithms in stochastic graphs. By modeling the infrastructure as a graph, potential vulnerabilities may be presented by computing what paths an attacker might take. This thesis project presents and compares two memory efficient algorithms that can be used to solve this problem approximately, the online k-means and sampled k-shortest path algorithm. By computing paths for several different graph types, the two algorithms were compared against a naive algorithm. The online k-means algorithm uses approximately 77 times less memory, executes in the same amount of time, and produces similar path lengths. In our experiments, the sampled k-shortest path algorithm uses approximately 154 times less memory, execution time is seen to be lowered by a factor of 5 to 20 depending on the graph type, but path distances computed are longer.

No description available.

The Diploma thesis deals with the problem of energy-efficient train control. It presents the basic survey of mathematical models used in the problem of energy-efficient train control, analysis of optimal driving regimes, determining optimal switching times between optimal driving regimes and timetabling of the train. The mathematical formulation of the problem is done using Newton's second law of motion and other known physical laws. To analyse optimal driving regimes and determine the switching times between optimal driving regimes, we apply tools of optimal control theory, particularly Pontryagin's Maximum Principle. The timetabling of the train is discussed from the numerical solution of the settled non-linear programming problem.

This diploma thesis deals with optimization models of financial risks. The first part, which is devoted to the theoretical background, introduces the basic concepts of optimization, modern portfolio theory, fundamental and technical analysis and statistical background. The basic principles of operation of modern portfolio theory are presented. The methods for analysis and selection of assets called Growth at A Reasonable Price and portfolio optimization approach according to Harry Markowitz were used with selected methods. The practical part is focused on the data analysis, selection of assets and design of a portfolio optimization model according to selected conditions with an emphasis on minimizing investment risk. The used models examine the selected data and are solved using the MS Excel add-in Solver version.

This diploma thesis addresses the economic assessment of the methods of heating and combinations of technologies for energy saving. The objective of this work is to assess the economical effectiveness of individual types of sources of heating. It is first necessary to established gross requirements of energy for heating, water heating, and other household operations. Based on this information, suitable sources are determined. Alternative energy sources are other aspects for consideration – controlled ventilation with heat recovery, and photovoltaic cells. The aim of the thesis is to identify which combination of those technologies is the most economically effective for the opted house. This work aims to help the investor in decision making about the technological solution of heating and power purchase. The outcome is a complex evaluation of different options for the assessed house.

This dissertation deals with the securities portfolio optimization. After introducing the definitions, I try to explain the particular investment instruments with regard to returns and risks. The following part provides a theory which tells more about different market risks and returns on the final securities portfolio. Concerning these models the effective portfolio has been set up.

abstract: The goal of reinforcement learning is to enable systems to autonomously solve tasks in the real world, even in the absence of prior data. To succeed in such situations, reinforcement learning algorithms collect new experience through interactions with the environment to further the learning process. The behaviour is optimized by maximizing a reward function, which assigns high numerical values to desired behaviours. Especially in robotics, such interactions with the environment are expensive in terms of the required execution time, human involvement, and mechanical degradation of the system itself. Therefore, this thesis aims to introduce sample-efficient reinforcement learning methods which are applicable to real-world settings and control tasks such as bimanual manipulation and locomotion. Sample efficiency is achieved through directed exploration, either by using dimensionality reduction or trajectory optimization methods. Finally, it is demonstrated how data-efficient reinforcement learning methods can be used to optimize the behaviour and morphology of robots at the same time. / Dissertation/Thesis / Doctoral Dissertation Computer Science 2019

Investors constantly seek to secure business ventures and structures that will provide them with the most tax-efficient consequences by utilising loopholes in tax legislation and exploiting them within the legal requirements. With the recent growing interest in the undeveloped markets in Africa, many South African companies aim to invest into Africa in a tax-efficient manner. Mauritius, being a low tax jurisdiction and having a favourable tax treaty network with a large number of African countries, is an attractive choice for South African companies wishing to set up a platform for investing into Africa. The aim of this study was to address the shortcomings of efficient tax planning and the approach to invest into Africa using Mauritius as gateway for South African resident companies. The study focused on the tax implications of an offshore trust and offshore company incorporated in Mauritius for tax-efficient investing in order to minimise taxes. Therefore this study did not focus on using Mauritius for tax evasion purposes and a qualitative approach was applied, using a hypothetical case study to determine the most tax-efficient organisational structure for minimising taxes. The findings of the study revealed that, on a balance of case law and tax legislation, a tax-minimising organisational structure is largely influenced by its residency status and South Africa‟s control foreign company (CFC) legislation. Residency for an offshore trust and offshore company will be at the place where it is effectively managed. The findings revealed that the tax consequences are similar for an offshore trust and offshore company in Mauritius legislation. However, the hypothetical case study revealed that the impact of the CFC legislation can have negative consequences for a structure where only an offshore company is used, and therefore the ideal tax-minimising structure will be where a South African company uses a combination of an offshore trust and offshore company in Mauritius in order to avoid the possibility of CFC legislation having an impact on such a structure. / Dissertation (MCom)--University of Pretoria, 2012. / Taxation / unrestricted

The present trend towards shorter product lives means that manufacturers are faced with less time whereby production costs can be controlled and reduced. As a result many companies are turning towards the use of Systems Engineering and cost management techniques that are intended to reduce product costs during the design phases of the product life. <p>One of the many cost management tools that are presently in use is target costing. An iterative process, target costing attempts to reduce the cost of manufacturing products by using value engineering techniques so that identified target costs can be realized. As a step improvement program, the ability of a company to obtain its target cost in not always certain. Target profits from individual products are grouped together into portfolios and then managed according to company strategic profit plans. <p>This report investigates methods used for identifying the uncertainties that can exist in the target costing process and suggests the potential use efficient portfolio techniques for the management of process oriented or continuous products. The report also suggests a number of follow-on projects that could encompass the use of specific decision support systems for the examination of both target costing uncertainties and portfolio management. / Master of Science