Orientador: Fernando Augusto Silva Marins / Resumo: Em diferentes aspectos da vida cotidiana, o ser humano é forçado a escolher entre várias opções, esse processo é conhecido como tomada de decisão. No nível do negócio, a tomada de decisões desempenha um papel muito importante, porque dessas decisões depende o sucesso ou o fracasso das organizações. No entanto, em muitos casos, tomar decisões erradas pode gerar grandes custos. Desta forma, alguns dos problemas de tomada de decisão que um gerente enfrenta comumente são, por exemplo, a decisão para determinar um preço, a decisão de comprar ou fabricar, em problemas de logística, problemas de armazenamento, etc. Por outro lado, a coleta de dados tornou-se uma vantagem competitiva, pois pode ser utilizada para análise e extração de resultados significativos por meio da aplicação de diversas técnicas, como estatística, simulação, matemática, econometria e técnicas atuais, como aprendizagem de máquina para a criação de modelos preditivos. Além disso, há evidências na literatura de que a criação de modelos com técnicas de aprendizagem de máquina têm um impacto positivo na indústria e em diferentes áreas de pesquisa. Nesse contexto, o presente trabalho propõe o desenvolvimento de um modelo preditivo para tomada de decisão, usando as técnicas supervisionadas de aprendizado de máquina, e combinando o modelo gerado com as restrições pertencentes ao processo de otimização. O objetivo da proposta é treinar um modelo matemático com dados históricos de um processo decisório e obter os predit... (Resumo completo, clicar acesso eletrônico abaixo) / Mestre

This master`s thesis analyzes the possibilities of investment business plan, which is related to the development of alternative energy sources in the Czech Republic. The main part consist an analysis, which lead to decision, if we start prepare or not in 2011 the project of renewable energy sources.

Air pollution harms the environment and human welfare. Computer models and their simulation are useful tools for deeper understanding of processes behind as they quite accurately represent the dispersion and transformation of pollutants with advection diffusion equation or by other concepts. Current models give valid results only to constrained cases of initial conditions. The general model combining the several specific models which is able to change according to input parametres and improve with training is proposed. The adaptiveness of the system is provided by decision tree as data structure with information for selection and combination process and genetic algorithm as optimization method for adjusting the tree. The evaluation of implemented system proves that the combination of models gives better results than models themselves. Even with simple specific models, the system has achieved results comparable to state-of-art models of air pollution.

This thesis deals with the extension of the bachelor's thesis Movie Database. I focus on the practical application of data mining as well as the personalization of this information system. In data mining, I will use the data accrued during the operation site. The introduction summarizes the concepts of data mining. In the second section  outlines new developments in web applications followed by an overview and description of the new site's features.  The last chapters are devoted to  the description of the design and the implementation of data mining techniques. The conclusion gives the achieved results.

This study investigates the predictive performance of two different machine learning (ML) models on the stock market and compare the results. The chosen models are based on artificial neural networks (ANN) and random forests (RF). The models are trained on two separate data sets and the predictions are made on the next day closing price. The input vectors of the models consist of 6 different financial indicators which are based on the closing prices of the past 5, 10 and 20 days. The performance evaluation are done by analyzing and comparing such values as the root mean squared error (RMSE) and mean average percentage error (MAPE) for the test period. Specific behavior in subsets of the test period is also analyzed to evaluate consistency of the models. The results showed that the ANN model performed better than the RF model as it throughout the test period had lower errors compared to the actual prices and thus overall made more accurate predictions. / Denna studie undersöker hur väl två olika modeller inom maskininlärning (ML) kan förutspå aktiemarknaden och jämför sedan resultaten av dessa. De valda modellerna baseras på artificiella neurala nätverk (ANN) samt random forests (RF). Modellerna tränas upp med två separata datamängder och prognoserna sker på nästföljande dags stängningskurs. Indatan för modellerna består av 6 olika finansiella nyckeltal som är baserade på stängningskursen för de senaste 5, 10 och 20 dagarna. Prestandan utvärderas genom att analysera och jämföra värden som root mean squared error (RMSE) samt mean average percentage error (MAPE) för testperioden. Även specifika trender i delmängder av testperioden undersöks för att utvärdera följdriktigheten av modellerna. Resultaten visade att ANN-modellen presterade bättre än RF-modellen då den sett över hela testperioden visade mindre fel jämfört med de faktiska värdena och gjorde därmed mer träffsäkra prognoser.

Recent developments in educational research raise important questions about the design of learning environments—questions that suggest the value of rethinking what is taught, how it is taught, and how is it assessed. During the past few decades, STEM disciplines began formally recognizing and integrating discipline-based education research (DBER) into their research programs to improve STEM education. One of the less literature-affluent areas of DBER addresses curriculum order and design appertaining to concept types and the order in which we teach those concepts. As educational researchers, we pose the question: does content order matter? In this project we designed, implemented and analyzed a concrete-to-abstract curriculum as a way of teaching and learning that not only builds off what students already know but how their intellect develops throughout the learning process. This semester-long curriculum design is scientifically supported and provides a learning environment aimed to not only building a student’s declarative knowledge of the subject but procedural knowledge as well and a way of developing scientific reasoning skills. This design also aimed at enhancing a student’s ability to make connections between biological concepts despite being classified as different biological concept types (e.g. descriptive, hypothetical, and theoretical concepts) as described by Lawson et al (2000). The reasoning behind and development of this project was based from Jean Piaget’s proposed stages of intellectual development, which supports the concrete-to-abstract theory. We found that, when compared to a traditional biology course (abstract-to-concrete in terms of content order), a concrete-to-abstract order of content resulted in significantly higher biological declarative knowledge and ability to make concept connections. While we failed to detect a significant difference between the two courses in terms of how quickly scientific reasoning skills are developed or how students’ scores on scientific reasoning skill assessments, the concrete-to-abstract course did show significantly higher gains in reasoning between the start and end of the semester. In addition to this project, a significant amount of time was also allocated to the design and evaluation of a health promotion and education program in Samoa. We developed a program which centered on a principal-run caregiver meeting as a means to expand health promotion and prevention efforts concerning Rheumatic Heart Disease, which is a significant cause of child morbidity and mortality in Samoa. We found that training principals on how to inform their student’s caregivers was an effective way to increase RHD awareness and disseminate correct health information including what to do if their child presents with a sore throat.

El presente trabajo tiene como finalidad aplicar los conocimientos y técnicas impartidas durante los tres cursos de Ciencia de Datos. Específicamente identificar y utilizar las variables encontradas en el negocio para determinar un modelo que permita una mayor permanencia del personal en la empresa Mister Pollo. En ese contexto, la investigación se apoyará en la metodología de ciencia de datos de IBM, la cual inicia con la fase de comprensión del negocio para identificar el problema de la organización, analizando sus fortalezas y debilidades; así como la fase de recopilación y preparación de los datos, análisis, interpretación, modelado y evaluación de la data. Asimismo, el tipo de investigación que se emplea es mixto, pues en la fase inicial tiene un enfoque descriptivo que permite entender la importancia de las variables utilizadas. En la segunda fase, el enfoque se vuelve predictivo gracias a la utilización de una técnica de aprendizaje supervisado, en este caso, el modelo de árbol de decisión para la determinación de una herramienta que permita evaluar la mayor permanencia de trabajadores en el restaurante. Esto permitirá que el Gerente General de la empresa pueda elaborar un plan de acción para poder controlar y minimizar la rotación del personal, considerando diferentes escenarios, perfiles y necesidades de la empresa. Finalmente, en la conclusión de este proyecto se evaluarán los hallazgos en el modelo seleccionado para verificar que responden a los objetivos planteados por el Gerente de la empresa Míster Pollo en coordinación con el equipo de trabajo. / The purpose of this work is to apply the knowledge and techniques taught during the three Data Science courses. Specifically, to identify and use the variables found in the business to determine a model that allows a greater permanence of the personnel in the company Mister Pollo. In this context, the research will be supported by IBM's data science methodology, which begins with the phase of understanding the business to identify the organization's problem, analyzing its strengths and weaknesses; as well as the phase of data collection and preparation, analysis, interpretation, modeling and evaluation of the data. Likewise, the type of research used is mixed, since in the initial phase it has a descriptive approach that allows understanding the importance of the variables used. In the second phase, the approach becomes predictive thanks to the use of a supervised learning technique, in this case, the decision tree model for the determination of a tool to evaluate the greater permanence of workers in the restaurant. This will allow the General Manager of the company to develop an action plan to control and minimize staff turnover, considering different scenarios, profiles and needs of the company. Finally, at the conclusion of this project, the findings of the selected model will be evaluated to verify that they respond to the objectives set by the Manager of the company Míster Pollo in coordination with the work team. / Trabajo de investigación

The master's thesis deals with the knowledge discover in databases and with the extending of the data mining systems in the Oracle environment developed at the VUT FIT. The system kernel conception incorporates an interface that enables the adding of data mining modules. The objective of the thesis is to learn this interface and implement and embed the data mining module for decision-tree classification into the application. In addition, the thesis compares the application with similar commercial product SAS Enterprise Miner

[pt] A metodologia Box & Jenkins tem sido mais utilizada para fazer previsões do que outros métodos até então. Alguns analistas têm relutado, entretanto, em usar esta metodologia, em parte porque a identificação da estrutura adequada é uma tarefa complexa. O reconhecimento tanto dos padrões de comportamento das funções de autocorrelação quanto da autocorrelação parcial (teórica/estimada) dependem da série temporal através da qual é possível extraí-las. Uma vez obtidos os resultados, pode-se inferir qual o tipo de estrutura Box & Jenkins adequada para a série. A proposta do trabalho é desenvolver três novas metodologias de identificação automática das estruturas Box & Jenkins ARMA simples e/ou sazonais, identificar os filtros sazonal e linear da série de uma forma menos complexa. A primeira metodologia utiliza árvores de decisão, a segunda, redes neurais e a terceira, K-Nearest Neighbor (KNN). A estas metodologias serão utilizadas as estruturas Box & Jenkins sazonais de períodos 3, 4, 6 e 12 e não sazonais. Os resultados são aplicados a séries simuladas, bem como a séries reais. Como comparação, utilizou-se o método automático de identificação proposto no software FPW-XE. / [en] The Box & Jenkins is the most popular forecasting technique. However, some researchers have not embraced it because the identification of its structure is highly complex. The process of proper characterizing the properties of both autocorrelation functions and partial correlation (theoretical or estimated) depends on the time series from which they are being obtained. Given the results in question, it is possible to infer the proper Box & Jenkins structure for the time series being studied. For the reasons above, the goal of this dissertation is to develop three new methodologies to identifying, in an automatic fashion, the Box & Jenkins structure of an ARMA series. The methodologies identify, in a simpler manner, both the seasonal and linear filters of the series. The first methodology applies the decision tree. The second applies the neural networks. The third applies the K-Nearest Neighbor (KNN). In each of them the Box & Jenkins seasonal structures of 3, 4, 6 and 12 periods were used, as well as the nonseasonal structure. The results are applied to simulated and actual series. For comparison purposes, the automatic identification procedure of the software FPW-XE is also used.

This thesis presents an activity mode intent recognition approach for safe, robust and reliable control of powered backbone exoskeleton. The thesis presents the background and a concept for a powered backbone exoskeleton that would work in parallel with a user. The necessary prerequisites for the thesis are presented, including the collection and processing of surface electromyography signals and inertial sensor data to recognize the user’s activity. The development of activity mode intent recognizer was described based on decision tree classification in order to leverage its computational efficiency. The intent recognizer is a high-level supervisory controller that belongs to a three-level control structure for a powered backbone exoskeleton. The recognizer uses surface electromyography and inertial signals as the input and CART (classification and regression tree) as the classifier. The experimental results indicate that the recognizer can extract the user’s intent with minimal delay. The approach achieves a low recognition error rate and a user-unperceived latency by using sliding overlapped analysis window. The approach shows great potential for future implementation on a prototype backbone exoskeleton.