The main objective of this study is to divide a number of colorectal cancer cases into subgroups based on their molecular features using cluster analysis. The data used is supplied by a research group at Pathology, the Department of Medical Biosciences, Umeå University, and consists, after some preparation, of 455 observations which is a larger data set than many similar studies. The molecular variables that the Clustering is based on are CIMP (CpG Island Methylator Phenotype), MSI (Micro Satellite Instability), BRAF- and KRAS-mutations. These are categorical variables and consequently the Clustering method used is PAM (Partitioning Around Medoids) which is particularly useful with data on diverse variable level. The final analysis results in four subgroups that are represented by different combinations of attributes on the aforementioned variables.

The main objective of this study is to divide a number of colorectal cancer cases into subgroups based on their molecular features using cluster analysis. The data used is supplied by a research group at Pathology, the Department of Medical Biosciences, Umeå University, and consists, after some preparation, of 455 observations which is a larger data set than many similar studies. The molecular variables that the Clustering is based on are CIMP (CpG Island Methylator Phenotype), MSI (Micro Satellite Instability), BRAF- and KRAS-mutations. These are categorical variables and consequently the Clustering method used is PAM (Partitioning Around Medoids) which is particularly useful with data on diverse variable level. The final analysis results in four subgroups that are represented by different combinations of attributes on the aforementioned variables.

The main objective of this study is to divide a number of colorectal cancer cases into subgroups based on their molecular features using cluster analysis. The data used is supplied by a research group at Pathology, the Department of Medical Biosciences, Umeå University, and consists, after some preparation, of 455 observations which is a larger data set than many similar studies. The molecular variables that the Clustering is based on are CIMP (CpG Island Methylator Phenotype), MSI (Micro Satellite Instability), BRAF- and KRAS-mutations. These are categorical variables and consequently the Clustering method used is PAM (Partitioning Around Medoids) which is particularly useful with data on diverse variable level. The final analysis results in four subgroups that are represented by different combinations of attributes on the aforementioned variables.

The main objective of this study is to divide a number of colorectal cancer cases into subgroups based on their molecular features using cluster analysis. The data used is supplied by a research group at Pathology, the Department of Medical Biosciences, Umeå University, and consists, after some preparation, of 455 observations which is a larger data set than many similar studies. The molecular variables that the Clustering is based on are CIMP (CpG Island Methylator Phenotype), MSI (Micro Satellite Instability), BRAF- and KRAS-mutations. These are categorical variables and consequently the Clustering method used is PAM (Partitioning Around Medoids) which is particularly useful with data on diverse variable level. The final analysis results in four subgroups that are represented by different combinations of attributes on the aforementioned variables.

SCA is the largest private owned forest owner in Sweden, with an ownership of 2 million hectares of productive forestland. To manage the planning of such large areas and to fit the traditional organization structure of a Swedish forest company, the planning approach is subdivided into three stages; strategic, tactical and operational. The strategic part, based on a statistic sample of stands, returns a harvest level estimate which should then be implemented at actual sites in the tactical part. This procedure can lead to incompatible solutions as the strategic solution is not necessarily possible to translate into actual sites in the tactical plan. The geographic aspect is important in tactical planning due to the large costs for moving equipment and maintenance of a large number of forest roads.

SCA is the largest private owned forest owner in Sweden, with an ownership of 2 million hectares of productive forestland. To manage the planning of such large areas and to fit the traditional organization structure of a Swedish forest company, the planning approach is subdivided into three stages; strategic, tactical and operational. The strategic part, based on a statistic sample of stands, returns a harvest level estimate which should then be implemented at actual sites in the tactical part. This procedure can lead to incompatible solutions as the strategic solution is not necessarily possible to translate into actual sites in the tactical plan. The geographic aspect is important in tactical planning due to the large costs for moving equipment and maintenance of a large number of forest roads.

I denna uppsats presenteras teknik för att automatiskt besvara frågor skrivna i naturligt språk, givet att man har tillgång till en samling tidigare ställda frågor och deras respektive svar.Jag bygger ett prototypsystem som utgår från en databas med epost-konversationer från HP Help Desk. Systemet kombinerar Latent Semantisk Analys med en täthetsbaserad klustringsalgoritm och en enkel klassificeringsalgoritm för att identifiera frekventa svar och besvara nya frågor.De automatgenererade svaren utvärderas automatiskt och resultaten jämförs med de som tidigare presenterats för samma datamängd. Inverkan av olika parametrar studeras också i detalj.Studien visar att detta tillvägagångssätt ger goda resultat, utan att man behöver utföra någon som helst lingvistisk förbearbetning..

The aim of this Master?s thesis is to investigate if the library can be a real partner for researchers in managing and dealing with their research information. In this study, six researchers have been interviewed about their problems with information management. This study uses the method Critical Incident Technique for the interviews with the researchers. This method focuses mainly on problems and what happens around those problems.

Denna rapport behandlar en undersökning av olika lastbalanseringsalgoritmer i Linux Virtual Server. Undersökningen har gjorts i ett webbkluster (Apache var webbservern som användes) med tre heterogena noder, där operativsystemet var den detalj som skiljde dem åt. Operativsystemen som ingick i undersökningen var Windows Server 2008 R2, CentOS 6.2 och FreeBSD 9.0. De faktorer som undersöktes mellan de olika algoritmerna var klustrets genomsnittliga svarstid vid olika belastning och hur många anslutningar som kunde hanteras av klustret, detta gjordes med verktyget httperf.Undersökningen ger svar på hur ett heterogent webbklusters genomsnittligasvarstid och arbetskapacitet kan skilja sig åt beroende på vilken algoritm som används för lastbalansering. Resultatet visar att den genomsnittliga svarstiden håller sig låg tills en hastig stigning inträffar.

In Economic Geography, there has been a strong interest of how different forms of proximity impacts businesses. In addition, there is also a strong interest in business agglomeration, as a result businesses cluster together through physical proximity, networks and knowledge. In the development sector, there is a strong interest on how to best support the creations and growth of small enterprises.This essay emphasises on the business cluster formations in Katutura, a suburb to Windhoek the capital city of Namibia. It oversees the impacts of different types of cluster formations and their effects on the businesses. In addition, this essay will identify the main clusters in Katutura and state what makes them unique from each other and what is a common cluster feature.

Inom alla verksamheter och organisationer finns det idag någon form av informationssystem. För att effektivisera, förändra och stödja verksamheten har allt fler företag valt att anskaffa ett standardsystem som hjälpmedel. Tyvärr innebär detta även att ett antal problem kan uppstå. Ett av de mest kritiska problemen som bör försöka undvikas är utebliven användaracceptans. Syftet med detta arbete var att undersöka vilka insatser som bör vidtas i arbetet med att anskaffa ett standardsystem, för att förbättra användaracceptansen.

Målet med den här studien är att undersöka hur barn under 15 år påverkades av Cryptosporidiumutbrottet i slutet av år 2010 i Östersunds kommun. Datamaterialet utgörs av svar på en enkätundersökning från 514 barn rörande deras hälsa relaterad till utbrottet. Dessa enkäter togs fram av svenska Smittskyddsinstitutet kort efter utbrottet och det är i uppdrag av denna myndighet som studien utförs. Analys av riskfaktorer och följdsymptom utförs med logistiska regressionsmodeller utifrån både ett Bayesianskt och ett frekventistiskt tillvägagångssätt för att på så sätt betrakta datamaterialet från fler synvinklar och samtidigt identifiera skillnader mellan dessa två tillvägagångssätten. En annan del av arbetet presenterar bortfallskalibrerade skattningar av antalet Cryptosporidiumfall både totalt och månadsvis men också skattningar av fallprevalensen i olika redovisningsgrupper.

The problem of separating a non-rectangular foreground image from a background image is a classical problem in image processing and analysis, known as matting or keying. A common example is a film frame where an actor is extracted from the background to later be placed on a different background. Compositing of these objects against a new background is one of the most common operations in the creation of visual effects. When the original background is of non-constant color the matting becomes an under determined problem, for which a unique solution cannot be found. This thesis describes a framework for computing mattes from images with backgrounds of non-constant color, using Bayesian statistics.

The Psychology Driving and Barriers to Skunk Work Projects (referred to throughout this thesis as the paper) represents a study predominantly on Tetra Pak from a selective perspective, the innovators. Innovators both past&present whose initiatives contribute to the well being of Tetra Pak as well as one from Ericsson were consulted. As with individuals, organizations possess definable characters that are influenced by the various stages of their life cycle. Theories of corporate life cycle phases and organizational character analysis were therefore employed to identify and explain organizational barriers to skunk work leaders&managers. Predicated on the Organizational Character Index survey developed by William Bridges PhD (himself a recognized leader in the field of transition management) and administered to sixty-one managers within a Tetra Pak business unit established the benchmark.

It is clear that media advertising is one of the crucial marketing tools for companies in order to promote their products and services. As we are moving towards more and more economic integration, both on a regional and global scale, the question about standardisation of media advertising has become increasingly important and is nowadays subject to discussion in most international companies. This thesis deals with the issue of standardisation in a region where the process of political, legal and especially economic integration is taking its course, but where cultural and linguistic diversity is still a remaining fact: Europe. The aim of this thesis was to investigate the potential gains from a standardised advertising approach but also to identify its limitations. Especially on a market place like Europe, with its diverse cultural and linguistic background, a fully standardised approach might be difficult to implement.