Análise de Dados e Sistemas de Apoio à Decisão / Master in Data Analysis and Decision Support Systems / Este estudo propõe a definição de uma metodologia de classificação que permite identificar as formações das equipas, no domínio do futebol robótico, na liga de simulação de duas dimensões (2D). Para alcançar esse objectivo foram utilizadas técnicas de Data Mining para problemas de classificação. Para explicar o funcionamento e características do futebol robótico simulado, com ênfase nos sistemas multi-agente, descreveram-se: a constituição do sistema de simulação de futebol com as respectivas regras, a comunicação entre o simulador e os jogadores e os respectivos protocolos, as percepções e acções dos agentes, os jogadores heterogéneos, o agente treinador, as suas funções e a sua linguagem de comunicação. Posteriormente, apresentaram-se as etapas do processo de Data Mining: preparação de dados, redução de dados, modelização e análise de solução. Neste trabalho a primeira etapa - preparação dos dados apresentou: a selecção das equipas de teste, a configuração do ambiente de simulação em Linux, a configuração da equipa FC Portugal, utilizada neste estudo, e o respectivo treino de forma a efectuar um jogo de futebol robótico simulado com dez formações distintas. Após a realização de seis jogos, utilizando quatro equipas distintas, procedeu-se à conversão dos log files desses jogos num conjunto de dados no formato típico (forma de matriz). Na segunda etapa efectuou-se a redução dos dados na escolha dos atributos de forma empírica, com base no conhecimento do processo de formações no futebol do mundo real e do futebol robótico simulado. Na modelização seleccionaram-se também, de forma empírica, os classificadores com potencialidade para produzirem o melhor modelo de previsão das formações elegendo-se para a análise da solução o indicador de avaliação taxa de erro, complementado com o teste estatístico t-Sudent, para amostras emparelhadas. Os resultados obtidos no conjunto de experiências realizadas demonstraram que é possível identificar com grande exactidão as formações utilizadas pela equipa FC Portugal em distintos jogos utilizando técnicas de Data Mining. Analisando os resultados foi possível concluir que os classificadores Sequential Minimal Optimization (SMO) e o k-Nearest Neighbor (IBK) obtiveram o melhor desempenho nas experiências realizadas. Finalmente concluiu-se ainda que o classificador mais indicado para gerar um modelo de previsão antes dos jogos de futebol robótico simulado é o SMO. / This study proposes a definition of one methodology of classification that let identify the formations of the teams, in domain of robotic soccer, in the simulation league of two dimensions (2D) league. To reach the goal of this study it was used techniques of Data Mining for classification problems. To explain the operation and the characteristics of robotic soccer simulated, with emphasis on multi-agent systems, is described: the constitution of the system simulation of soccer (football) with the respective rules, the communication between the simulator and the players and the respective protocols, the perceptions and agents actions, the heterogeneous players, the coach agent, their functions and their language of communication. Posteriorly, is presented the stages of Data Mining process: data preparation, data reduction, modeling and solution analysis, In this work the first stage data preparation presented: the selection of the test teams, the configuration of the simulation environment in Linux, the configuration of FC Portugal team, used in this study, and their training in order to make a game of robotic soccer simulated with ten different formations. After the completion of the six games, using four distinct teams was made the conversion of the log files, of these games, in a dataset with the typical format (matrix form). In the second stage was carried out the data reduction of the attributes in the empirical way, based on the knowledge of formations process in the real world soccer and in the robotic soccer simulated. In modeling were selected too in the empirical way, the classifiers with potential to produce the best forecast model of the formations. In the stage for solution analysis, the main indicators for evaluation were the error rate and the statistical test t-Student for paired samples. The results in the set of experiments demonstrated that it was possible to identify, with great accuracy, the formations used by the team FC Portugal in distinct games using techniques of Data Mining. Analysing the results it is possible to deduce that the classifiers Sequential Minimal Optimization (SMO) and the k-Nearest Neighbor (IBK) obtained the best performance in the experiments performed. Finally it was concluded that the most appropriate classifier to generate a forecast model before the games in robotic soccer simulated is the SMO.

In my work I deal with the issue of data mining and its use in the commercial sphere. Specifically, I focused on marketing and sales forecast. The aim of my work was first to assemble knowledge from data mining and then use it to create sales forecasts using data mining add-on for Excel. In the first part I gather the theoretical information about data mining. I focused on definition, the methodology, algorithms and of course on the most frequent usage. The second part consists of the practical application of acquired knowledge. I focus on making the sales forecast of HERO CZECH company. I used data mining add-on for Microsoft Excel 2007. Results are compared with real forecasts prepared by the Key Account Manager. Results of my work proved that forecasts from data mining add-on for Excel were not more accurately than the existing ones from Key Account Manager. Nevertheless, I believe that the use of data mining methods have found use in preparing the forecast, at least as a means of support.

This thesis explores the use of spatiotemporal data mining in the air quality domain to understand causes of PM2.5 air pollution. PM2.5 refers to fine particulate matter less than 2.5 microns in diameter and is a major threat to human and environmental health. A review of air quality modeling methods is provided, emphasizing data-driven modeling techniques. While data mining methods have been applied to air quality data, including temporal sequence mining algorithms, spatiotemporal sequence mining methods have not been broadly applied to study air pollution. However, air pollution is highly spatial in nature, so such methods can offer new insights into air quality. This thesis applies one such method, the Spatiotemporal Sequence Miner (STS Miner) algorithm, to air quality data from a low-cost sensor network to explore causes and trends related to PM2.5. To facilitate the use of this method, an open-source library called OpenSTSMiner is developed to implement this algorithm. Various domain results are found; for instance, low temperature and low relative humidity are strongly associated with worsening levels of air quality. Lastly, to highlight the utility of the STS Miner algorithm, a comparison is presented between STS Miner and spatial Markov chains, another spatiotemporal modeling method used in the air quality domain.

The rapid growth of e-commerce contributes to not only an increase in the number of online shoppers but also new changes in customer behaviour. Surveys have revealed that online shopper's brand loyalty and store loyalty are declining. Also the transparency of feedback affects customers' purchase intention. In the context of these changes, online sellers are faced with challenges in regard to their customer relationship managements (CRM). They are interested in identifying high-value customers from a mass of online shoppers, and knowing the factors that might have impacts on those high-value customers. This thesis aims to address these questions. Our research is conducted based on an eBay dataset that includes transaction and associated feedback information during the second quarter of 2013. Focusing on the sellers and buyers in that dataset, we propose an approach for measuring the value for each seller-buyer pair so as to help sellers capture high-value customers. For a seller, the value of each of its customers has been obtained, and we create a customer value distribution for the seller so that the seller knows the majority of its customers' consumption abilities. Next, we categorize sellers based on their customer value distributions into four different groups, representing the majority of customers as being of high, medium, low, and balanced values, respectively. After this classification, we compare the performance of each group in terms of the sales, percentage of successful transactions, and the seller level labelled by the eBay system. Furthermore, we perform logistic regression and clustering to the sellers' feedback data in order to investigate whether a seller's reputation has an impact on the seller's customer value distribution. From the experiment results, we conclude that the effect of negative ratings is more significant than that of positive ratings on a seller's customer value distribution. Also higher ratings about "Item as Described" and "Shipping and Handling Charges" are more likely to help the seller attract more high-value buyers. / Graduate

ITC/USA 2011 Conference Proceedings / The Forty-Seventh Annual International Telemetering Conference and Technical Exhibition / October 24-27, 2011 / Bally's Las Vegas, Las Vegas, Nevada / This paper presents a case study of an FTI system with complex requirements in terms of the data acquisition, recording, and post-analysis. Gigabit Ethernet was the technology of choice to facilitate such a system. Recording in a Gigabit Ethernet environment raises a fresh challenge to perform fast data reduction and data mining for post-flight analysis. This paper describes the Quick Access Recorder used in this system and how it addresses this challenge.

published_or_final_version / Business / Doctoral / Doctor of Philosophy

published_or_final_version / abstract / Mathematics / Doctoral / Doctor of Philosophy

Artificial Intelligence Lab, Department of MIS, University of Arizona / This bookâ s purpose is to present a balanced and integrated view of what a Knowledge Management System (KMS) is. We first define Knowledge Management (KM) from various consulting and IT perspectives and then pay particular attention to new and emerging technologies that help promote this new field. In particular, we present a review of some key KMS sub-fields: search engines, data mining, and text mining. We hope to help readers better understand the emerging technologies behind knowledge management, i.e., Knowledge Management Systems. A high-level, although systematic, discussion of text mining is presented. Unlike search engines and data mining that have a longer history and are better understood, text mining is an emerging technical area that is relatively unknown to IT professionals. We therefore present several case studies and conclude with lessons learned and future research and development directions. This book is intended to provide a gentle introduction to researchers and IT professionals who are new to KMS. We hope it provides a non-technical and practical review of this fascinating field as well as a look at the potential and pitfalls of these emerging technologies.

Based off Nicholson's 2000 University of North Texas dissertation, "CREATING A CRITERION-BASED INFORMATION AGENT THROUGH DATA MINING FOR AUTOMATED IDENTIFICATION OF SCHOLARLY RESEARCH ON THE WORLD WIDE WEB" located at http://scottnicholson.com/scholastic/finaldiss.doc / This research creates an intelligent agent for automated collection development in a digital library setting. It uses a predictive model based on facets of each Web page to select scholarly works. The criteria came from the academic library selection literature, and a Delphi study was used to refine the list to 41 criteria. A Perl program was designed to analyze a Web page for each criterion and applied to a large collection of scholarly and non-scholarly Web pages. Bibliomining, or data mining for libraries, was then used to create different classification models. Four techniques were used: logistic regression, non-parametric discriminant analysis, classification trees, and neural networks. Accuracy and return were used to judge the effectiveness of each model on test datasets. In addition, a set of problematic pages that were difficult to classify because of their similarity to scholarly research was gathered and classified using the models. The resulting models could be used in the selection process to automatically create a digital library of Web-based scholarly research works. In addition, the technique can be extended to create a digital library of any type of structured electronic information.

An integrated graph-theoretic and geometric approach to the analysis of aqueous solvation of atomic ions is presented. This analysis makes use of a novel data-mining tool, moleculaRnetworks, to process data from molecular dynamics simulations. The workings and structure of this tool are discussed, along with the development and testing of its PageRank algorithm-based rapid solvation polyhedra classifier. The ability to classify instantaneous solvation polyhedra enables a finely detailed understanding of shell structure-behavior relationships, as water molecules simultaneously rearrange about ions, exchange with the bulk, and rearrange their hydrogen-bond network. The application of the tool to cation systems, including lithium, sodium, potassium, magnesium, calcium, and lanthanum, yields new insight into the mechanisms of water exchange about these ions. It is shown that in order for exchange events to occur, the solvation shell must "preorganize" to admit or expel a molecule of water: this preorganization is reflected in the mechanistic preference for each ion. The application of the tool to anion systems, including fluoride, chloride, and bromide, reveals that these ions have an extended effect on the reorientation ability of water molecules beyond their first solvation shell. Finally, when both ions are present, as in the potential of mean force simulation between lanthanum and chloride, structural rearrangements can be seen as the ions break through the barrier to form the contact ion pair. Taken together, these results show the utility of the moleculaRnetworks tool in broadening our understanding of aqueous ion solvation.