Syftet med denna kandidatuppsats är att undersöka om och hur det går att kunna förutspå antalet gäster för en specifik skidanläggning i Sverige. Eftersom skidanläggningar är dyra att bedriva är det en viktig aspekt att kunna planera personal kostnadseffektivt. Genom att analysera skidortens stora datamängder angående historiska kunddata, tillsammans med historiska och reala väderdata, kan prediktiva analyser genomföras. Detta leder till att skidorten kan utforma bättre tillsättning av personal för att reducera liftköer i backarna, minska matsvinnet i restauranger och även minska eventuella förluster kopplade till överbemanning. Tack vare system som framkallar beslutsunderlag, så kallade beslutsstödsystem, kan företag agera konkurrenskraftigt på marknaderna. Den här studien försöker därför undersöka huruvida det går att framkalla en eventuell prognos för framtida gästantal. Genom att samla in olika typer av både kund- respektive väderdata, har tvättning av data genomförts för att sedan låta olika prediktiva modeller förutspå framtiden. Resultatet för studien påvisar betydelsen gällande bearbetningsprocessen av data, och avslutas med intressanta tankar gällande framtida forskning. Utifrån detta kan det konstateras att en eventuell prediktion är möjlig, men endast i mån av en ungefärlig gräns utifrån antalet gäster. Ett överskridande av gränsen riskerar prediktionsförmågan att försämras.

Det finns många olika Data Mining-processer som kan tillämpas i ett Data Mining-projekt. Fördelen med att använda en Data Mining-process är att projektet blir strukturerat, processen kan hjälpa till att minska risker som annars kan uppstå och kan medföra att projektmålet förändras. Data Mining-processen som studien har undersökt är generell och studien försöker därmed precisera olika faser av processen, för att anpassas till ett prognostiseringsprojekt.Studien utvärderar den preciserade prognostiseringsprocessen genom att följa och dokumentera ett prognostiseringsprojekt på en tillverkningsindustri. Studien analyserar teoretiskt vilka implikationer tillverkningsindustrin kan möta och även om studiens framtagna process är tillämpningsbar i detta fall. Studien visar att det på en teoretisk nivå går att genomföra studiens preciserade Data Mining-process och visar även vilka risker som kan uppkomma om ett prognostiseringsprojekt inte följer en Data Mining-process. / Program: Systemarkitekturutbildningen

<p>Syftet med detta arbete har varit att ta reda på elevers kunskaper om de geometriska begreppen dm³ och m² samt hur lärarna undervisar i dessa avsnitt. Detta har vi tagit fram genom att först göra en litteraturstudie följt av intervjuer. </p><p>Av litteraturen fick vi fram att laborativt arbete och materiel är essentiella delar i geometriundervisningen och har så varit under hela skolans historia. Det gäller att använda materielen på rätt sätt för att få ett bra resultat.</p><p>Resultatdelen, med intervjuer, visade att elevernas kunskaper gällande begreppen är vaga och ofullständiga. Deras resultat var sämre än vad vi trodde med undantag för de starka eleverna. Lärarna ger underbetyg till sin egen undervisning när de misstänker att eleverna har en dålig uppfattning om begreppen. Lärarna uppvisar varierande grad av laborativt arbete som till största delen är tillfredsställande och de ställer sig till största delen positivt till att arbeta laborativt.</p>

Syftet med detta arbete har varit att ta reda på elevers kunskaper om de geometriska begreppen dm3 och m2 samt hur lärarna undervisar i dessa avsnitt. Detta har vi tagit fram genom att först göra en litteraturstudie följt av intervjuer. Av litteraturen fick vi fram att laborativt arbete och materiel är essentiella delar i geometriundervisningen och har så varit under hela skolans historia. Det gäller att använda materielen på rätt sätt för att få ett bra resultat. Resultatdelen, med intervjuer, visade att elevernas kunskaper gällande begreppen är vaga och ofullständiga. Deras resultat var sämre än vad vi trodde med undantag för de starka eleverna. Lärarna ger underbetyg till sin egen undervisning när de misstänker att eleverna har en dålig uppfattning om begreppen. Lärarna uppvisar varierande grad av laborativt arbete som till största delen är tillfredsställande och de ställer sig till största delen positivt till att arbeta laborativt.

Para llevar a cabo en forma sistemática el proceso de descubrimiento de conocimiento en bases de datos, conocido como minería de datos, es necesaria la implementación de una metodología. Actualmente las metodologías para minería de datos se encuentran en etapas tempranas de madurez, aunque algunas como CRISP-DM ya están siendo utilizadas exitosamente por los equipos de trabajo para la gestión de sus proyectos. En este trabajo se establece un análisis comparativo entre las metodologías de minería de datos más difundidas en la actualidad. Para lograr dicha tarea, y como aporte de esta tesis, se ha propuesto un marco comparativo que explicita las características que se deberían tener en cuenta al momento de efectuar esta confrontación.

RNA localization, a regulated step of gene expression, is fundamentally important in development and differentiation. In multidisciplinary experiments, we discovered that RNA (mis)localization underlies the human disease myotonic dystrophy (DM). DM, the most prevalent adult muscular dystrophy, is caused independently by two alleles: DM1 is characterized by a (CTG)n expansion in the DM kinase (DMPK) gene 3' untranslated region while DM2 has a mutation in a small presumptive RNA binding protein. These analyses were guided by disease characteristics and have provided insights to DM's cytopathology, cell biology and molecular genetics. Examining muscle biopsies, it is demonstrated here that DM kinase mRNA is specifically subcellularly localized within normal human muscle and that DM kinase mRNA harboring the 3’UTR mutation (DM1) is mislocalized in DM patient muscle to cytoplasmic areas characteristic of DM disease pathology. Thus, the disease mutation alters the cellular distribution of the effected message. DMPK mRNA mislocalization causes altered DM kinase protein localization, correlates with novel phosphoprotein appearance and can account for DM’s diseased phenotype. While we were fortunate to access DM patient tissue to establish these key findings, the system does not lend itself to experimental manipulation. Hence, I established a disease- relevant tissue culture system, which recapitulates DMPK trafficking, Employing this system; I elucidate a complementary role for the DM2 gene product as a localization factor for DMPK mRNA (DM1 gene product). Comprehensive RNA-protein interaction experiments reveal the DM2 protein specifically and selectively recognizes a small, definitive area within the DMPK RNA 3'UTR. Detailed biochemical, cytological and functional experiments reveal 1) the DM2 protein colocalizes with DMPK mRNA, 2) the small area of the DMPK 3’UTR bound by pDM2 acts to properly localize a reporter construct and 3) disruption of the DM2 protein results in DMPK mRNA mislocalization. These data establish mRNA localization as a vital process underlying human disease etiology. Moreover, they reveal DM1 and DM2 gene products function in the same molecular pathway and that mutation of either causes DMPK mRNA mislocalization, leading to disease. These data have apparent application to several neuromuscular disorders and open a plethora of novel research avenues, both basic and applied. / text

Dissertação (mestrado)—Universidade de Brasília, Instituto de Ciências Exatas, Departamento de Ciência da Computação, 2008. / Submitted by Jaqueline Oliveira (jaqueoliveiram@gmail.com) on 2008-11-28T12:09:59Z No. of bitstreams: 1 DISSERTACAO_2008_GilbertoOliveiraHiragi.pdf: 895106 bytes, checksum: 181e2a9a782456ffe4637fc5519e09c8 (MD5) / Approved for entry into archive by Georgia Fernandes(georgia@bce.unb.br) on 2009-02-11T16:01:26Z (GMT) No. of bitstreams: 1 DISSERTACAO_2008_GilbertoOliveiraHiragi.pdf: 895106 bytes, checksum: 181e2a9a782456ffe4637fc5519e09c8 (MD5) / Made available in DSpace on 2009-02-11T16:01:26Z (GMT). No. of bitstreams: 1 DISSERTACAO_2008_GilbertoOliveiraHiragi.pdf: 895106 bytes, checksum: 181e2a9a782456ffe4637fc5519e09c8 (MD5) / Os bancos de germoplasma do SIBRARGEN (Sistema Brasileiro de Informações em Recursos Genéticos) funcionam como um grande catálogo das espécies vegetais e de seus acessos (tipos característicos dentro de um grupo ou variabilidades dentro da espécie), contendo mais de 100 mil acessos catalogados. Esses bancos incluem a identificação do acesso (passaporte), descrição dos aspectos genótipos (caracterização) e descrição dos aspectos fenótipos (avaliação) e permitem aos pesquisadores dessa área realizarem consultas SQL mas recuperando apenas os dados armazenados, resultantes da resolução das expressões booleanas utilizadas como critérios de busca. Essas consultas não facilitam a descoberta de novos conhecimentos ou a construção de modelos de previsão ou descrição. Essa pesquisa propõe uma metodologia de mineração de dados, derivada do modelo de referência CRISP/DM, que auxilie a exploração dessas bases de dados por pesquisadores não vinculados à área de informática (por exemplo, biólogos ou agrônomos) visando facilitar a realização de tarefas previstas nas seguintes fases do CRISP/DM: entendimento do negócio, compreensão dos dados, preparação de dados, modelagem, avaliação dos modelos gerados e colocação em uso. Para materializar a metodologia proposta e automatizar a sua utilização por parte de não-informatas, foi implementada a ferramenta HaDog (Hiragi Approach for Data Mining of Germoplasm). HaDog foi implementada utilizando a linguagem Java, banco de dados Oracle® versão 10g release 2 e é acessível através de uma interface Web, disponível aos pesquisadores credenciados para acesso ao SIBRARGEN. A metodologia de mineração de germoplasma proposta foi avaliada de forma experimental através de dois estudos de casos conduzidos com o apoio de pesquisadores da Embrapa Recursos Genéticos e Biotecnologia: determinação de acessos representativos de uma espécie ou grupo de espécies e proposição de coletas direcionadas, ambos problemas típicos de interesse do curador (pesquisador responsável pelo banco de germoplasma de uma espécie). Essa avaliação experimental mostrou que é possível introduzir os especialistas na área na utilização de técnicas de mineração de dados na base de germoplasma sem requerem que eles se envolvam em atividades de programação. Os resultados experimentais obtidos até o momento demonstram que o HaDog pode se constituir em um importante facilitador para a mineração das bases do SIBRARGEN, visando, principalmente, a descoberta de novos conhecimentos pelos especialistas. _________________________________________________________________________________________ ABSTRACT / The banks of germplasm of the SIBRARGEN (Brazilian Information System in Genetic Resources) function as a great catalogue of the vegetal species and of its accesses (characteristic types inside of a group or variabilities inside of the species), contend more than 100 thousand catalogued accesses. These banks include the identification of the access (passport), description of the genotypes aspects (characterization) and phenotype description (evaluation) and allow researchers of this area to carry through SQL queries but recouping only the stored data, resultant of the resolution of the used boolean expressions as criteria search. These queries don’t facilitate to the discovery of new knowledge or the construction of forecast models or description. This research considers a data mining methodology, derived from the model of reference CRISP/DM, that assists the exploration of these databases for researchers tied with the computer science area (for example, biologists or agronomists) aiming to facilitate the accomplishment of tasks foreseen in the following phases of the CRISP/DM: business understanding, data understanding, data preparation, modeling, evaluation of the generated models and deployment. To materialize the methodology proposal and to automatize its use by people who aren’t of the computer science area, the HaDog tool was implemented (Hiragi Approach of Data Mining of Germplasm). HaDog was implemented using the Java language, database Oracle® version 10g release 2 and is accessible through a Web interface, available to the credential researchers for access to the SIBRARGEN. The methodology of mining of germplasm proposal was evaluated of experimental form through two studies of cases lead with the support of researchers of the Embrapa (Genetic Resources and Biotechnology: determination of representative accesses of a species or group of species and proposal of directed collections, both typical problems of interest of the custodian (responsible researcher for the Bank of germplasm of a species). This experimental evaluation showed that it is possible to introduce the specialists in the area in the use of techniques of mining of data in the base of germplasm without require that they become involved themselves in activities of programming. The experimental results obtained so far show that HaDog can be a major facilitator for the mining of foundations of SIBRARGEN, targeting mainly, the discovery of new knowledge by specialists.

Final thesis deals with information-mining from large sets of medical data using methods and machine learning algorithms. The subject of the theoretical part is machine learning and its distribution, description of the basic data types in data mining, most important classifications and predictions methods, criterion defining the quality of prediction methods, description of data mining methodology and frequently used systems. The practical part focuses on statistical and informatics survey of provided medical data, appropriate transformation, subsequent design and implementation of experiments using machine learning methods to acquire new knowledge and hidden information and finally interpretation of the results together with conclusions for target groups.

abstract: The dawn of Internet of Things (IoT) has opened the opportunity for mainstream adoption of machine learning analytics. However, most research in machine learning has focused on discovery of new algorithms or fine-tuning the performance of existing algorithms. Little exists on the process of taking an algorithm from the lab-environment into the real-world, culminating in sustained value. Real-world applications are typically characterized by dynamic non-stationary systems with requirements around feasibility, stability and maintainability. Not much has been done to establish standards around the unique analytics demands of real-world scenarios. This research explores the problem of the why so few of the published algorithms enter production and furthermore, fewer end up generating sustained value. The dissertation proposes a ‘Design for Deployment’ (DFD) framework to successfully build machine learning analytics so they can be deployed to generate sustained value. The framework emphasizes and elaborates the often neglected but immensely important latter steps of an analytics process: ‘Evaluation’ and ‘Deployment’. A representative evaluation framework is proposed that incorporates the temporal-shifts and dynamism of real-world scenarios. Additionally, the recommended infrastructure allows analytics projects to pivot rapidly when a particular venture does not materialize. Deployment needs and apprehensions of the industry are identified and gaps addressed through a 4-step process for sustainable deployment. Lastly, the need for analytics as a functional area (like finance and IT) is identified to maximize the return on machine-learning deployment. The framework and process is demonstrated in semiconductor manufacturing – it is highly complex process involving hundreds of optical, electrical, chemical, mechanical, thermal, electrochemical and software processes which makes it a highly dynamic non-stationary system. Due to the 24/7 uptime requirements in manufacturing, high-reliability and fail-safe are a must. Moreover, the ever growing volumes mean that the system must be highly scalable. Lastly, due to the high cost of change, sustained value proposition is a must for any proposed changes. Hence the context is ideal to explore the issues involved. The enterprise use-cases are used to demonstrate the robustness of the framework in addressing challenges encountered in the end-to-end process of productizing machine learning analytics in dynamic read-world scenarios. / Dissertation/Thesis / Doctoral Dissertation Computer Science 2016

Attaining training objectives is the measure of a successful training as objectives defines the purpose of instructional events. Application of the training objectives is challenging in large and complex military trainings. The trainings in military domain not only focus on the completion of the trainings but effectively achieving the objectives of the training is the goal of the exercises. It has been realized that the performance to achieve the goal is strengthen by the instructional processes and materials which are crafted to address specific training objectives. Simulation is one of the effective and realistic learning tools which can be used in trainings. As it is known that simulation generates enormous data, analysis of this data which may contain hidden information is a challenging task. The use of data mining is a solution to this problem. The aim of this project is to propose a framework of a system for the instructors which can be followed for evaluating trainee’s performance so that their fulfillment of the training objectives can be improved. A proposal which is studied in this project is learning from previous training experiences using data mining techniques to improve the effectiveness of the training by predicting the performance of the trainee. For selecting the good prediction model to estimate the learning outcome of the trainees, different classification techniques have been compared. CRISP-DM model is considered as a base for proposing the framework in this dissertation. Proposed framework is then applied on the dataset obtained from the Swedish Military for the exercises which involved shooting the target.