39449 research outputs found
Sort by
Occupational exposure to bioaerosols in the Norwegian salmon processing industry
No full textObjectives Workers in salmon processing plants are at risk of respiratory diseases. The aim of this study was to describe the Norwegian salmon processing industry in respect to production-related factors that may influence the generation of bioaerosols in the work atmosphere, and to assess salmon processing workers’ personal exposure to protein and endotoxin. Methods The study comprised 222 workers from 9 plants. Fullshift personal exposure measurements of total protein (inhalable aerosol fraction, n = 380) and endotoxin (total aerosol sampler, n = 178) were collected on 4 consecutive workdays. Technical and process-related information was collected through plant visits and meetings with technical and production staff. Linear mixed-effect model was used, treating individuals as random effect and work area and work task within areas as fixed effects. Results Plants differed in size, setup, processing procedures, and use of labor along the processing lines. Salmon processing overall geometric mean (GM) exposure to inhalable protein across the plants was highest in filleting area with 4.83 µg/m3 (geometric standard deviation [GSD] 3.16), followed by 3.91 µg/m3 (GSD 2.42) in slaughtering area, and 1.68 µg/m3 (GSD 2.40) in other areas. Endotoxin levels were generally low with the highest levels in slaughtering (GM 0.24 EU/m3; GSD 3.48), followed by other area (GM 0.19 EU/m3; GSD 4.05) and filleting (GM 0.10 EU/m3; GSD 2.51). The overall correlation between inhalable protein and endotoxin (total aerosol sampler) was poor (r = 0.13, P = 0.12). Conclusions Salmon processing workers are exposed to airborne inhalable protein bioaerosols at levels similar to those measured over a decade ago, indicating that a systematic approach to reduce exposure levels is still needed. Given the known health risk, the industry and regulatory bodies need to intensify efforts to reduce exposure and protect workers’ health. The variance in exposure levels to inhalable protein across plants, areas, and tasks might form the basis for better exposure-reducing strategies.publishedVersio
The Introduction of Chickens Into Norway and Their Early Use: New Evidence From the Borgund Kaupang
No full textChickens reached areas of northern Europe by the 6th to 5th century bce, but their dispersal into Scandinavia appears delayed.Here we present a thorough assessment of chicken remains recovered from Borgund, a deserted late Viking Age–Medieval urbansite located near Ålesund on the west coast of Norway. Direct and indirect dating of 20 chicken bones gives an age range fromthe Viking Age to the boundary between the High and Late Medieval. The Borgund chickens thus include some of the earliestevidence for chicken-keeping in Norway. The absence of juveniles indicates that chickens were kept for secondary products, suchas eggs and feathers, and societal reasons rather than just meat. This is in line with data on chickens from Medieval Norway. Thelow percentage of chickens in comparison to other domestic species indicates chickens were not a vital part of the day-to-day dietof the people of Borgund. The chickens here represent the earliest unambiguous record for the west coast and second earliest forNorway, indicating a Viking Age introduction. The existence of Viking Age trading networks with northwestern Norway andDenmark suggests that chickens may have been introduced to Borgund through this route. Alternatively, chickens may havebeen first introduced in southern Norway via a separate Viking Age network and then spread from there. The route throughwhich chickens came to Borgund remains unclear.publishedVersio
Reward Design and Self-Play for Deep Reinforcement Learning in Hydropower Scheduling
No full textMixed Integer Linear Programming and Stochastic Dual Dynamic Programming are
traditional methods used to solve the hydropower scheduling problem. However, they are
limited by computational complexity and scalability. This thesis explores the application
of deep reinforcement learning (DRL) as a potential solution to these limitations,
specifically using the Proximal Policy Optimization (PPO) algorithm. Two different
reward functions, Rest-price-Aware and Importance-Weighted Reward, are constructed and
evaluated on convergence speed, policy behavior and economic performance. Additionally,
self-play, a reinforcement learning approach where agents are trained by competing
against themselves is introduced and tested to mitigate the reliance on manually crafted
reward functions. Experimental results from short-horizon scenarios (5-time-steps) show
that all reward functions achieve near optimal performance when compared to the optimal
solution. Extended-horizon experiments (14-time-steps) reveal larger discrepancies
between the agents, revealing the importance of the reward function design. Furthermore,
self-play demonstrated strong potential by learning a versatile policy without relying on
a handcrafted reward function, and even outperformed the agents trained with manually
crafted reward functions. The findings suggest the potential of DRL in hydropower
scheduling, particularly approaches that limit reliance prior domain knowledge, could
serve as a valuable complement to existing hydropower scheduling methods and warrant
further investigation in more realistic and complex settings.Masteroppgave i informatikkINF399MAMN-PROGMAMN-IN
Terror and Policy: Why Some Attacks Reshape Laws and Others Fade Away
No full textAcross the globe, many countries have to deal with the threat and aftermath of terrorist attacks. While some countries rapidly change laws as a result of individual attacks to give the population a sense of security, others are struck by numerous attacks every year and yet do not consider it necessary to introduce any policy changes. Consequently, not all terrorist attacks are the same – they can differ in the place and time at which they occur, the motivation for the attack, the means by which they are carried out, the possible number of victims and any other damage they may cause. This thesis, therefore, examines the question as to why only some terrorist attacks lead to changes in legislation while others do not. Based on a qualitative comparative analysis (QCA) of a total of 107 cases of terrorist attacks in EU member states between the years 2006 and 2023, the causal complexity of policy-making in times of focusing events is addressed. Four pathways for legislative change are established, which are primarily guided by religious motivation, their weapon of choice, and a high number of victims. One of these combinations is contextualised and visualised by comparing it to a terrorist attack included in the dataset – that of November 13, 2015, in Paris. The analysis shows a connection between the motivation for an attack and the number of victims to the resulting policy outcomes, whereby other factors have an influence only if they contribute to higher victim numbers.Master's Thesis in Politics and Governance of Global ChallengesGOV380MASV-GLGO
Multiple Loads Moving on Ice Sheets
No full textThis master’s thesis serves as a continuation of the work done in K. Johnsen, “Wave response to an arbitrary motion of a load on an ice plate,” University of Bergen, 2022, where a solution for arbitrary pressure distributions acting on an ice sheet was presented. The derivation of this problem and a corrected solution is presented, and the solution is applied to multi-load scenarios. The exact solution depends on an integral, and is expressed in Fourier-space, thus numerical calculations are used to evaluate the solution of the problem under different conditions. The main purpose of my work into this problem was writing code that can be used to make these computations. This code is given and can also be found at https://github.com/ThomasJakobsenCodes/Multiple-Loads-Moving-on-Ice-Sheets .Masteroppgave i anvendt og beregningsorientert matematikkMAB399MAMN-MA
Likhetshensynet ved rettede emisjoner i noterte allmennaksjeselskap: bør regelverket endres for å styrke likebehandling av aksjonærer?
No full textFinanstilsynets gjennomgang av emisjonspraksisen blant noterte selskaper i Norge viste at det som etter loven skal være et unntak, i praksis er den ubestridelige hovedregel. Andelen rettede emisjoner er svært høy, også i EØS-sammenheng. Dette ga grunn til å spørre om aksjonærene i noterte selskap er sårbar for usaklig forskjellsbehandling. Med bakgrunn i det, analyserer denne masteroppgaven om det er behov for en endring av regelverket for å bedre sikre likebehandling, og derigjennom vareta tilliten til det norske kapitalmarkedet.MasteroppgaveJUS399MAJUR-2MAJU
Combinatorial optimization problems and solution methods for offshore monitoring and maintenance.
No full textDenne avhandlingen undersøker optimaliseringsbaserte tilnærminger for planlegging og beslutningstaking i komplekse offshore-operasjoner, med særlig fokus på miljøovervåking og vedlikehold av havvindparker. Disse områdene innebærer betydelige logistiske utfordringer, inkludert begrensede ressurser og behovet for å koordinere flere aktører, som fartøy, autonome kjøretøy og menneskelige teknikere. Effektiv planlegging i slike settinger er avgjørende for å fremme bærekraft, redusere kostnader og øke driftssikkerheten.
For å møte disse utfordringene utvikler og analyserer avhandlingen en rekke modeller og algoritmer for ruteplanlegging, dekningsproblemer og tidsplanlegging. Et sentralt tema er konseptet om varierende dekning, der området som kan overvåkes fra en posisjon avhenger dynamisk av hvor lenge man oppholder seg der. Dette er spesielt relevant for overvåking i havområder, hvor fysiske prosesser som havstrømmer kan føre forurensning langt fra kilden. Basert på dette utforsker forskningen både én- og flerfartøysystemer, inkludert scenarier med koordinering mellom et bemannet fartøy og en flåte av autonome undervannsfarkoster.
Metodisk kombinerer arbeidet matematisk programmering, heuristikker og maskinlæringsteknikker, særlig dyp forsterkende læring, for å utvikle adaptive løsningsstrategier som kan håndtere skalaen og kompleksiteten i virkelige problemstillinger. Metodene testes gjennom omfattende beregningseksperimenter og anvendes i praktiske casestudier, inkludert overvåking av offshore CO-lagringssteder og vedlikeholdsplanlegging for havvindparker.
Samlet bidrar avhandlingen med nye problemformuleringer, løsningsmetoder og innsikter i skjæringspunktet mellom operasjonsanalyse og intelligente systemer, med anvendelser som støtter bærekraftige og effektive offshore-operasjoner.This thesis investigates optimization-based approaches for planning and decision-making in complex offshore operations, with a focus on environmental monitoring and the maintenance of offshore wind farms. These domains present significant logistical challenges, including limited resources and the need to coordinate multiple agents such as vessels, autonomous vehicles, and human technicians. Efficient planning in these settings is crucial for improving sustainability, reducing costs, and increasing operational reliability.
To address these challenges, the thesis develops and analyzes a range of models and algorithms for routing, coverage, and scheduling problems. A central theme is the concept of varying coverage, where the area that can be monitored from a location depends dynamically on the time spent there. This idea is particularly relevant for ocean monitoring, where physical processes such as currents can carry pollutants away from their source. Building on this, the research explores both single-vehicle and multi-vehicle systems, including scenarios involving coordination between a manned vessel and a fleet of autonomous underwater vehicles.
Methodologically, the work combines mathematical programming, heuristics, and machine learning techniques, notably deep reinforcement learning, to develop adaptive solution strategies that can handle the scale and complexity of real-world instances. These methods are tested through extensive computational experiments and applied to practical case studies, including marine monitoring of offshore CO storage sites and maintenance planning for offshore wind farms.
Collectively, this thesis contributes new problem formulations, solution methods, and insights at the intersection of operations research and intelligent systems, with applications that support sustainable and efficient offshore operations.Doktorgradsavhandlin
Keeping silent in criminal cases - when silence speaks loudest
No full textMasteroppgaveJUS399MAJUR-2MAJU
Voices of Doubt: A Qualitative Study of Climate Skepticism Among Norwegian Political Parties and the Public
No full textThis thesis examines climate skepticism in Norway by comparing its expression among selected
political parties and the public. The study examines whether the skepticism expressed by
political parties aligns with that found in public opinion. The analysis focuses on four parties:
the Progress Party (Fremskrittspartiet), the Norwegian Democrats (Norgesdemokratene), the
Industry and Business Party (Industri- og Næringspartiet), and the People’s Party (Folkets
parti). The study identifies how different forms of skepticism are articulated through qualitative
content analysis of party manifestos and semi-structured interviews with party representatives.
Public attitudes are examined using data from the Norwegian Citizen Panel and CICERO
reports.
Climate skepticism is categorized into five primary types: trend, attribution, impact, response,
and epistemic skepticism. Additionally, anti-establishment rhetoric is included as an analytical
dimension. While the Progress Party expresses moderate skepticism, primarily questioning the
effectiveness of climate policy, the Norwegian Democrats and the Industry and Business Party
adopt more radical positions, rejecting both the scientific consensus and the legitimacy of
national and international climate governance. The People’s Party primarily criticizes symbolic
measures and international agreements. In contrast, most Norwegians accept that climate
change is real and largely human-induced. Attributional skepticism exists, particularly among
certain demographic groups, but trend skepticism is rare. The study finds partial alignment
between party rhetoric and specific population segments but also reveals that some parties adopt
more extreme or oppositional positions than the public at large. These findings contribute to a
broader understanding of how climate skepticism functions within political discourse and its
implications for climate policy legitimacy in Norway.Master's Thesis in Politics and Governance of Global ChallengesGOV380MASV-GLGO
PP-rådgiverens yrkesutøvelse og handlingsrom i praksis - En kvalitativ studie av opplevelser erfaringer
No full textDenne masteroppgaven undersøker hvordan rådgivere i Pedagogisk-Psykologisk tjeneste (PPT) opplever handlingsrommet i sin daglige yrkesutøvelse. Problemstillingen er knyttet til å forstå hvilke faktorer som påvirker rådgivernes handlingsrom. Videre søker studien å belyse hvordan PP-rådgivere balanserer organisatoriske rammer, politiske føringer og egen autonomi i et komplekst arbeidsmiljø.
Studien har et kvalitativt design, basert på semistrukturerte intervjuer med fire PP-rådgivere fra ulike kommunale kontorer. Intervju er valgt som metode for å vise rådgivernes egne erfaringer og opplevelse av handlingsrom og hvilke faktorer de erfarer at påvirker deres yrkesutøvelse. Intervjumaterialet er analysert og inspirert av Malterud sin forståelse av kvalitativ analyse for å identifisere sentrale mønstre og variasjoner i materialet.
Oppgavens teoretiske rammeverk tar utgangspunkt i Lipsky sin teori om bakkebyråkratiet, som beskriver hvordan yrkesutøvere i første linje forvalter sitt handlingsrom innenfor stramme rammer, og ofte i skjæringspunktet mellom motstridende krav. I tillegg brukes Etzioni sin teori om semi-profesjoner til å belyse kompleksiteten i PP-rådgivernes yrkesrolle.
Funnene viste at PP-rådgivere ser det som viktig å være nært praksisfeltet, og har et ønske om å hjelpe. Samtidig kommer det frem at de kan oppleve yrkesrollen ulikt. Særlig fremstår ressurstildeling og muligheter til å medvirke i egen arbeidshverdag som viktige faktorer for rådgivernes opplevde handlingsrom. Opplevelsen av handlingsrom synes også å være påvirket av rådgivernes utdanningsbakgrunn, etiske holdninger og organisatoriske rammer. Et annet interessant funn er hvordan krysspresset mellom individrettet arbeid og økende forventninger om systemarbeid påvirker deres opplevelser i daglig praksis. Videre er et interessant funn at kollegial støtte trekkes frem som viktig for å håndtere utfordringer og utøve faglig autonomi.Masteroppgave i spesialpedagogikkSPED395MAPS-SPEDMAPS-PE