El transporte de carga por medio terrestre es una de las actividades con altos números de accidentes e incidentes en la actualidad debido a los diversos factores que influyen en esta tarea. Si a esto se le añade el transporte de materiales peligrosos como los hidrocarburos (combustible), explosivos y lubricantes la probabilidad de que se maximice el evento es elevado y perjudicial para todos los actores externos e internos de la actividad. En el Perú existen leyes y normativas que regulan el transporte de materiales peligrosos, así como también el cumplimiento total de cada una de estas. Por ello, que las empresas brindan las capacitaciones y condiciones a los trabajadores que desempeñan esta labor, sin embargo, existe un factor elemental para evitar la ocurrencia de accidentes/incidentes la cual está arraigada al comportamiento del trabajador; así como también a la motivación y concientización de realizar sus actividades de forma segura. En este contexto, el presente trabajo de investigación plantea la implementación del Coaching en el Programa de Seguridad Basada en el Comportamiento (SBC) para disminuir los accidentes e incidentes, así como también para aumentar las conductas seguras de los trabajadores dentro de una empresa de transporte de materiales peligrosos. Este proyecto ejecutará el diseño y la aplicación del Coaching en el área de operaciones que se encuentra relacionada directamente al transporte de los materiales peligrosos. / Freight transportation by land is one of the activities with high numbers of accidents and incidents today due to the various factors that influence this task. If to this is added the transport of hazardous materials such as hydrocarbons (fuel), explosives and lubricants, the probability that the event will be maximized is high and detrimental to all the external and internal actors of the activity. In Peru there are laws and regulations that regulate the transport of hazardous materials, as well as the full compliance with each of these. Therefore, the companies provide training and conditions to workers who perform this work, however, there is an elementary factor to avoid the occurrence of accidents / incidents which is rooted in the behavior of the worker; as well as the motivation and awareness to carry out their activities safely. In this context, this research work proposes the implementation of Coaching in the Behavior Based Safety Program (SBC) to reduce accidents and incidents, as well as to increase the safe behaviors of workers within a transportation company of hazardous materials. This project will execute the design and application of Coaching in the area of operations that will be directly related to the transport of hazardous materials. / Tesis

This thesis studies the Chinese banks’ credit risk assessment using the Post Keynesian approach. We argue that bank loans are the major financial sources in emerging economies and it is uncertainty, an unquantifiable risk, rather than asymmetric information about quantifiable risk, as held by the mainstream approach, which is most important for the risk attached to credit loans, and this uncertainty is particularly important in China. With the universal existence of uncertainty, borrowers and lenders have to make decisions based on convention and experience. With regard to the nature of decision-making, this implies the importance of qualitative methods rather than quantitative methods. The current striking problem in Chinese banking is the large amount of Non-Performing Loans (NPLs) and this research aims to address the NPLs through improving credit risk management. Rather than the previous literature where Western models are introduced into China directly or with minor modification, this work advocates building on China’s conventional domestic methods to deal with uncertainty. We briefly review the background of the Chinese banking history with an evolutionary view and examine Chinese conventions in the development of the credit market. Based on an overview of this history, it is argued that Soft Budget Constraints (SBC) and the underdeveloped risk-assessing mechanism contributed to the accumulation of NPLs. Informed by Western models and experience, we have made several suggestions about rebuilding the Chinese convention of credit risk assessment, based on an analysis of publications and interviews with Chinese bankers. We also suggest some further development of the Asset Management Companies (AMCs) which are used to dispose of the NPLs.

In this thesis Generalized Additive Neural Networks (GANNs) are studied in the context of predictive Data Mining. A GANN is a novel neural network implementation of a Generalized Additive Model. Originally GANNs were constructed interactively by considering partial residual plots. This methodology involves subjective human judgment, is time consuming, and can result in suboptimal results. The newly developed automated construction algorithm solves these difficulties by performing model selection based on an objective model selection criterion. Partial residual plots are only utilized after the best model is found to gain insight into the relationships between inputs and the target. Models are organized in a search tree with a greedy search procedure that identifies good models in a relatively short time. The automated construction algorithm, implemented in the powerful SAS® language, is nontrivial, effective, and comparable to other model selection methodologies found in the literature. This implementation, which is called AutoGANN, has a simple, intuitive, and user-friendly interface. The AutoGANN system is further extended with an approximation to Bayesian Model Averaging. This technique accounts for uncertainty about the variables that must be included in the model and uncertainty about the model structure. Model averaging utilizes in-sample model selection criteria and creates a combined model with better predictive ability than using any single model. In the field of Credit Scoring, the standard theory of scorecard building is not tampered with, but a pre-processing step is introduced to arrive at a more accurate scorecard that discriminates better between good and bad applicants. The pre-processing step exploits GANN models to achieve significant reductions in marginal and cumulative bad rates. The time it takes to develop a scorecard may be reduced by utilizing the automated construction algorithm. / Thesis (Ph.D. (Computer Science))--North-West University, Potchefstroom Campus, 2006.

In this thesis Generalized Additive Neural Networks (GANNs) are studied in the context of predictive Data Mining. A GANN is a novel neural network implementation of a Generalized Additive Model. Originally GANNs were constructed interactively by considering partial residual plots. This methodology involves subjective human judgment, is time consuming, and can result in suboptimal results. The newly developed automated construction algorithm solves these difficulties by performing model selection based on an objective model selection criterion. Partial residual plots are only utilized after the best model is found to gain insight into the relationships between inputs and the target. Models are organized in a search tree with a greedy search procedure that identifies good models in a relatively short time. The automated construction algorithm, implemented in the powerful SAS® language, is nontrivial, effective, and comparable to other model selection methodologies found in the literature. This implementation, which is called AutoGANN, has a simple, intuitive, and user-friendly interface. The AutoGANN system is further extended with an approximation to Bayesian Model Averaging. This technique accounts for uncertainty about the variables that must be included in the model and uncertainty about the model structure. Model averaging utilizes in-sample model selection criteria and creates a combined model with better predictive ability than using any single model. In the field of Credit Scoring, the standard theory of scorecard building is not tampered with, but a pre-processing step is introduced to arrive at a more accurate scorecard that discriminates better between good and bad applicants. The pre-processing step exploits GANN models to achieve significant reductions in marginal and cumulative bad rates. The time it takes to develop a scorecard may be reduced by utilizing the automated construction algorithm. / Thesis (Ph.D. (Computer Science))--North-West University, Potchefstroom Campus, 2006.

Vérification Formelle et Incrémentale de Spécifications SysML pour la Conception de Systèmes à Base de ComposantsLe travail présenté dans cette thèse est une contribution à la spécification et la vérification des Systèmes à Base de Composants (SBC) modélisé avec le langage SysML. Les SBC sont largement utilisés dans le domaine industrielet ils sont construits en assemblant différents composants réutilisables, permettant ainsi le développement de systèmes complexes en réduisant leur coût de développement. Malgré le succès de l'utilisation des SBC, leur conception est une étape de plus en plus complexe qui nécessite la mise en {\oe}uvre d'approches plus rigoureuses.Pour faciliter la communication entre les différentes parties impliquées dans le développement d'un SBC, un des langages largement utilisé est SysML, qui permet de modéliser, en plus de la structure et le comportement du système, aussi ses exigences. Il offre un standard de modélisation, spécification et documentation de systèmes, dans lequel il est possible de développer un système, partant d'un niveau abstrait, vers des niveaux plus détaillés pouvant aboutir à une implémentation. %Généralement ces systèmes sont faits plus grands parce qu'ils sont développés avec des cadres logiciels.Dans ce contexte nous avons traité principalement deux problématiques.La première est liée au développement par raffinement d'un SBC modélisé uniquement par ses interfaces SysML. Notre contribution permet au concepteur des SBC de garantir formellement qu'une composition d'un ensemble de composants élémentaires et réutilisables raffine une spécification abstraite d'un SBC. Dans cette contribution, nous exploitons les outils: Ptolemy pour la vérification de la compatibilité des composants assemblés, et l'outil MIO Workbench pour la vérification du raffinementLa deuxième problématique traitée concerne la difficulté de déterminer quoi construire et comment le construire, en considérant seulement les exigences du système et des composants réutilisables, donc la question qui en découle est la suivante: comment spécifier une architecture SBC qui satisfait toutes les exigences du système? Nous proposons une approche de vérification formelle incrémentale basée sur des modèles SysML et des automates d'interface pour guider, par les exigences, le concepteur SBC afin de définir une architecture de système cohérente, qui satisfait toutes les exigences SysML proposées. Dans cette approche nous exploitons le model-checker SPIN et la LTL pour spécifier et vérifier les exigences.Mots clés: {Modélisation, Spécifications SysML, Architecture SBC, Raffinement, Compatibilité, Exigences, Propriétés LTL, Promela/SPIN, Ptolemy, MIO Workbench} / Formal and Incremental Verification of SysML Specifications for the Design of Component-Based SystemsThe work presented in this thesis is a contribution to the specification and verification of Component-Based Systems (CBS) modeled in SysML. CBS are widely used on the industrial field, and they are built by assembling various reusable components, allowing developing complex systems at lower cost.Despite the success of the use of CBS, their design is an increasingly complex step that requires the implementation of more rigorous approaches.To ease the communication between the various stakeholders in a CBS development project, one of the widely used modeling languages is SysML, which besides allowing modeling of structure and behavior, it has capabilities to model requirements. It offers a standard for modeling, specifying and documenting systems, wherein it is possible to develop a system, starting from an abstract level, to more detailed levels that may lead to an implementation.In this context, we have dealt mainly two issues. The first one concerns the development by refinement of a CBS, which is described only by its SysML interfaces and behavior protocols. Our contribution allows the designer of CBS to formally ensure that a composition of a set of elementary and reusable components refines an abstract specification of a CBS. In this contribution, we use the tools: Ptolemy for the verification of compatibility of the assembled components and MIO Workbench for refinement verification.The second one concerns the difficulty to decide what to build and how to build it, considering only system requirements and reusable components. Therefore, the question that arises is: how to specify a CBS architecture, which satisfies all system requirements? We propose a formal and incremental verification approach based on SysML models and interface automata to guide, by the requirements, the CBS designer to define a coherent system architecture that satisfies all proposed SysML requirements. In this approach we use the SPIN model-checker and LTL properties to specify and verify requirements.Keywords: {Modeling, SysML specifications, CBS architecture, Refinement, Compatibility, Requirements, LTL properties, Promela/SPIN, Ptolemy, MIO Workbench}

In this dissertation, generalized additive neural networks (GANNs) and multilayer perceptrons (MLPs) are studied and compared as prediction techniques. MLPs are the most widely used type of artificial neural network (ANN), but are considered black boxes with regard to interpretability. There is currently no simple a priori method to determine the number of hidden neurons in each of the hidden layers of ANNs. Guidelines exist that are either heuristic or based on simulations that are derived from limited experiments. A modified version of the neural network construction with cross–validation samples (N2C2S) algorithm is therefore implemented and utilized to construct good MLP models. This algorithm enables the comparison with GANN models. GANNs are a relatively new type of ANN, based on the generalized additive model. The architecture of a GANN is less complex compared to MLPs and results can be interpreted with a graphical method, called the partial residual plot. A GANN consists of an input layer where each of the input nodes has its own MLP with one hidden layer. Originally, GANNs were constructed by interpreting partial residual plots. This method is time consuming and subjective, which may lead to the creation of suboptimal models. Consequently, an automated construction algorithm for GANNs was created and implemented in the SAS R statistical language. This system was called AutoGANN and is used to create good GANN models. A number of experiments are conducted on five publicly available data sets to gain insight into the similarities and differences between GANN and MLP models. The data sets include regression and classification tasks. In–sample model selection with the SBC model selection criterion and out–of–sample model selection with the average validation error as model selection criterion are performed. The models created are compared in terms of predictive accuracy, model complexity, comprehensibility, ease of construction and utility. The results show that the choice of model is highly dependent on the problem, as no single model always outperforms the other in terms of predictive accuracy. GANNs may be suggested for problems where interpretability of the results is important. The time taken to construct good MLP models by the modified N2C2S algorithm may be shorter than the time to build good GANN models by the automated construction algorithm / Thesis (M.Sc. (Computer Science))--North-West University, Potchefstroom Campus, 2011.

In this dissertation, generalized additive neural networks (GANNs) and multilayer perceptrons (MLPs) are studied and compared as prediction techniques. MLPs are the most widely used type of artificial neural network (ANN), but are considered black boxes with regard to interpretability. There is currently no simple a priori method to determine the number of hidden neurons in each of the hidden layers of ANNs. Guidelines exist that are either heuristic or based on simulations that are derived from limited experiments. A modified version of the neural network construction with cross–validation samples (N2C2S) algorithm is therefore implemented and utilized to construct good MLP models. This algorithm enables the comparison with GANN models. GANNs are a relatively new type of ANN, based on the generalized additive model. The architecture of a GANN is less complex compared to MLPs and results can be interpreted with a graphical method, called the partial residual plot. A GANN consists of an input layer where each of the input nodes has its own MLP with one hidden layer. Originally, GANNs were constructed by interpreting partial residual plots. This method is time consuming and subjective, which may lead to the creation of suboptimal models. Consequently, an automated construction algorithm for GANNs was created and implemented in the SAS R statistical language. This system was called AutoGANN and is used to create good GANN models. A number of experiments are conducted on five publicly available data sets to gain insight into the similarities and differences between GANN and MLP models. The data sets include regression and classification tasks. In–sample model selection with the SBC model selection criterion and out–of–sample model selection with the average validation error as model selection criterion are performed. The models created are compared in terms of predictive accuracy, model complexity, comprehensibility, ease of construction and utility. The results show that the choice of model is highly dependent on the problem, as no single model always outperforms the other in terms of predictive accuracy. GANNs may be suggested for problems where interpretability of the results is important. The time taken to construct good MLP models by the modified N2C2S algorithm may be shorter than the time to build good GANN models by the automated construction algorithm / Thesis (M.Sc. (Computer Science))--North-West University, Potchefstroom Campus, 2011.

No description available.

During the last century, various countries' armed forces have used unmanned aerial vehicles, commonly known as drones. In recent years, strives have been made to develop small commercial drones that have allowed the general public to afford and use them for recreational purposes. The availability of drones has led to immoral applications of the technology. Such applications need to be faced with anti-measures and effective detection methods. Therefore, this thesis aims to develop a mobile reconnaissance robot that can detect commercial drones with radar. It describes integrating radar sensors with single-board computers to detect and localise air-bound objects. The finished product aims to be used for educational and exhibition purposes at the Swedish Armed Forces technical school to increase awareness of the technology. / <p>Försvarsmaktens tekniska skola i Halmstad var intressenter för uppsatsen.</p>

No description available.