This study examines the wastewater treatment system on BillerudKorsnäs integrated pulpand paper mill at Karlsborg, with focus on the effluent of suspended solids. Due to lowerregulatory limits for the emissions of suspended solids BillerudKorsnäs is looking forways to improve the efficiency of the wastewater treatment. Therefore, this study seeksto create a statistical model for the emissions of suspended solids and to find which factorsin the production, wastewater treatment and surrounding environment that affects theemissions. The purpose of the study is to extend the current understanding of the treatmentsystem to create a base for future decisions. The methods used for the modelling areElastic net regression and Random forest classifier, which were selected for the variableselection properties they provide. Unfortunately, the results show that the fitted modelscan only explain a small part of the variation of the emissions of suspended solids.The lack of fit in the models indicate that the current data is not sufficient to explain thevariation in the emissions of suspended solids. During the study deficiencies in the datacollection has been detected and improvements to overcome these deficiencies are proposedin the hope of creating more reliable data for the company and to create a betterbasis for future studies. / Denna studie undersöker avloppsvattenreningen på BillerudKorsnäs integrerade massaoch pappersbruket i Karlsborg, med fokus på utsläppen av suspenderade ämnen. Pågrund ut av sänkta gränser för utsläppen av suspenderade ämnen undersöker BillerudKorsnäsmöjligheter för att öka effektiviteten i avloppsvattenreningen. Därför försöker dennastudie att skapa en statistisk modell för utsläppen av suspenderade ämnen och undersökavilka faktorer i produktion, avloppsvattenrening, och omgivande miljö som påverkar utsläppen.Syftet med studien är att förbättra förståelsen för avloppsvattenreningen och attskapa beslutsunderlag för framtida förbättringar. Metoderna som används i modellernaär Elastic net regression och Random forest classifier, vilka valdes för deras variabel selectionegenskaper. Resultaten för modellerna visar att modellerna endast kan förklara enliten del av variationen i utsläpp av suspenderade ämnen. Det dåliga resultatet indikeraratt nuvarande data inte är tillräckligt för att skapa den typ av modeller som används. Understudien så har brister i datainsamlingen upptäckts, och förbättringar för att överbryggadessa brister är föreslagna med ändamål att göra insamlad data mer pålitlig och skapa enbättre bas för framtida studier.

One of the most important fields regarding growth and development for mostorganizations today is the digitalization, or digital transformation. The offering oftechnological solutions to enhance existing, or create new, processes or products isemerging. That is, it’s of great importance that organizations continuously affirm thepotential of applying new technical solutions into their existing processes. For example, a well implemented AI solution for automation of an existing process is likely tocontribute with considerable business value.Medical underwriting for individual insurances, which is the process consideredin this project, is all about risk assessment based on the individuals medical record.Such task appears well suited for automation by a machine learning based applicationand would thereby contribute with substantial business value. However, to make aproper replacement of a manual decision making process, no important informationmight be excluded, which becomes rather challenging due to the fact that a considerable fraction of the information the medical records consists of unstructured textdata. In addition, the underwriting process is extremely sensible to mistakes regarding unnecessarily approve insurances where an enhanced risk of future claims can beassessed.Three algorithms, Logistic Regression, XGBoost and a Deep Learning model, wereevaluated on training data consisting of the medical records structured data from categorical and numerical answers, the text data as TF-IDF observation vectors, and acombination of both subsets of features. The XGBoost were the classifier performingbest according to the key metric, a pAUC over an FPR from 0 to 0.03.There is no question about the substantial importance of not to disregard anytype of information from the medical records when developing machine learning classifiers to predict the medical underwriting outcomes. At a very risk conservative andperformance pessimistic approach the best performing classifier did manage, if consider only the group of youngest kids (50% of sample), to recall close to 50% of allstandard risk applications at a false positive rate of 2%, when both structured andtext data were considered. Even though the structured data accounts for most of theexplanatory ability it becomes clear that the inclusive of the text data as TF-IDF observation vectors make for the differences needed to potentially generate a positivenet present value to an implementation of the model

This paper addresses the question of whether Russia’s role conception changed after the annexation of Crimea. Research on role conceptions has a solid standing within foreign policy analysis, but information on role change and what it is that affects it has not been given the necessary attention. The paper’s theoretical framework contains earlier theoretical definitions of role conceptions and role change. In conjunction with the theoretical approach, process tracing and systematic text analysis with a focus on narratives are used as methodological tools. Relevant findings are then generated through the study of elite-expressions of the Russian master role from, firstly, the period between 2012 and late 2013, and, secondly, 2014. In brief, the results indicate that no major change occurred. Rather the role conception remained stable throughout both sequences. This suggests that the ramifications of the annexation of Crimea did not challenge the Russian master role and that it was not perceived as a crisis by Putin and his colleagues. Moreover, it is possible that role change more likely occurs when exceptional circumstances arise unexpectedly and when they fundamentally challenge ingrained role conceptions.

No description available.

This study investigates resistance against female local government politicians in Sri Lanka during their first year and a half as members of local government councils, as a result of the 25 % gender quota introduced in 2018. Further, the study investigates these newly elected female politicians’ perceived ability to influence local government politics; experienced substantive representation. Through a minor field study, in-depth interviews were conducted to examine forms of resistance and perceived political influence of these newly appointed women. Several types of resistance were found and divided into three categories; Patronizing Behavior from Male Politicians; the Dispute Between Elected and Appointed Women; Public Distrust. Some evidence of the mandate effect and the label effect were detected, further hampering substantive representation. The study concludes that despite a significant increase in descriptive representation, substantive representation was not necessarily experienced by the interview subjects. Further studies are encouraged to deepen the understanding of the resistance towards appointed female politicians in Sri Lanka, and moreover the problematic effects of the implementation of gender quotas in highly unequal states.

The terrestrial carbon cycle is a part of the global carbon cycle, where one important component is the terrestrial vegetation. Terrestrial vegetation largely controls the land surface carbon exchanges and leverage the atmospheric greenhouse gas concentrations, significantly affecting the trajectory of global warming. It is therefore important to improve the understanding of vegetation response to different climatic factors, in particular for those linked to large-scale climate variability, which is still less studied so far. Vegetation greenness is suggested to be a useful tool in order to understand vegetation response. Looking at Europe, the climate factors that affect vegetation the most are linked to the large-scale atmospheric circulation over the North Atlantic, like the jet stream, which varies in speed and latitude, and the North Atlantic Oscillation (NAO). Here, I compute monthly indices representing the variability of these atmospheric features, and correlate them with monthly vegetation greenness data (NDVI) anomalies over a period of five years. This is done both for regionally-averaged NDVI and the months April-July and as a geographical point-by-point analysis for the month of May. The results show a significant correlation between Scandinavian NDVI and the NAO as well as jet speed at multiple time lags, up until 2 months. The jet latitude, instead, showed significant correlation for three regions in mid/southwestern Europe at longer time lags of 3-4 months. This means that the position of the jet in winter can affect the spring vegetation growth in this area. The jet speed and NAO, however, works mostly at shorter timespans. / Den jordbunda kolcykeln, som är en del av den globala kolcykeln, består av olika komponenter där en viktig del är vegetation. Växtlighet på land kontrollerar till stor del utbytet av kol vid jordytan och har därigenom inflytande på atmosfäriska växthusgaskoncentrationer, vilket medför stor påverkan på global uppvärmning. Det är därför viktigt att förbättra förståelsen för hur vegetation reagerar på olika klimatologiska faktorer, särskilt de som är kopplade till storskalig klimatvariabilitet då dessa kopplingar har studerats i mindre utsträckling hittils. Ett bra sätt att mäta den jordbunda kolcyklen på är med grönhet av vegatation. Om vi beaktar Europa så är det främst storskaliga atmosfäriska cirkulatoiner över norra Atlanten av de klimatologiska faktorerna som påverkar vegetation. En av dessa faktorer är jetströmmen, vilken varierar i fart och latitud, samt Nordatlantiska Oscillationen(NAO). I detta arbete beräknar jag index som representerar variationen i dessa i form av månadsgenomsnitt och korrelerar dem med månatlig data över avvikelser i vegetationsgrönhet (NDVI) över en femårsperiod. Det här gjordes för både regionala medelvärden och månaderna april-juli samt en geografisk punkt till punkt analys utförd för maj. Resultatet visar att det finns en signifikant korrelation mellan NDVI i Skandinavien och NAO samt jetfarten vid flera tidsfördröjningar, upp till 2 månader. Jetlatituden visade däremot signifikant korrelation för tre regioner i centrala/sydvästa Europa vid längre tidsfördröjningar på 3-4 månader. Detta innebär att positionen på jetströmmen under vintern kan påverka vegetationstillväxten under våren i detta område. Jetfarten och NAO påverkar däremot mest vid kortare tidsspan.

Frequentist model averaging has started to grow in popularity, and it is considered a good alternative to model selection. It has recently been applied favourably to gen- eralized linear models, where it has mainly been purposed to aid the prediction of probabilities. The performance of averaging estimators has largely been compared to that of models selected using AIC or BIC, without much discussion of model screening. In this paper, we study the performance of model averaging in classification problems, and evaluate performances with reference to a single prediction model tuned using cross-validation. We discuss the concept of model screening and suggest two methods of constructing a candidate model set; averaging over the models that make up the LASSO regularization path, and the so called LASSO-GLM hybrid. By means of a Monte Carlo simulation study, we conclude that model averaging does not necessarily offer any improvement in classification rates. In terms of risk, however, we see that both methods of model screening are efficient, and their errors are more stable than those achieved by the cross-validated model of comparison.

The main objective of this study is to analyse problem representations within national and multilateral policy concerning environmental migration in Central America. The study mainly focuses on Mexico and Costa Rica’s national legal frameworks regarding environmental migration but also draws on bilateral as well as multilateral agreements ratified by the countries. In a two-step analysis, the perspectives of human security and state security are used to identify key representations, followed by an application of Bacchi’s (2016) post-structural policy analysis tool ‘What is the problem represented to be?’, allowing for an understanding of environmental migration policy in a wider context. The results of the study show how human security characteristics are most prevalent within environmental migration policy, albeit acknowledging the implicit prevalence of state security characteristics. The study makes attributions to the understanding of the discourse and conceptualisation concerning environmental migration and recommends further studies on efficient interlinkages between human and state security-oriented policies.

No description available.

The 2018 Swedish general election saw the largest collective polling error so far in the twenty-first century. As in most other advanced democracies Swedish pollsters have faced extensive challenges in the form of declining response rates. To deal with this problem a new method based on search query data is proposed. This thesis predicts the Swedish general election using Google Trends data by introducing three models based on the assumption, that during the pre-election period actual voters of one party are searching for that party on Google. The results indicate that a model that exploits information about searches close to the election is in general a good predictor. However, I argue that this has more to do with the underlying weight this model is based on and little to do with Google Trends data. However, more analysis needs to be done before any direct conclusion, about the use of search query data in election prediction, can be drawn.