NHH Brage (Norges Handelshøyskole)
Not a member yet
8813 research outputs found
Sort by
What is fair? Assessing fairness of hospital networks in Norway: A study of the Innlandet region
Get PDFThere has been significant research related to fairness concerns in the healthcare space, from philosophical questions asking how societies should distribute resources, to papers addressing who healthcare decision-makers should be: the government, healthcare professionals, the public, or other stakeholders. This paper collects theories from distributive justice, procedural justice, and public healthcare decision-making to explore fair, yet feasible, alternatives to best find healthcare solutions for policy makers, in the context of mathematical modeling with optimization.
The Innlandet case study reveals the difficulty of coming to a concrete answer to these questions as the 20-year debate, removing two hospitals in Hamar and Gjøvik municipalities in favor of a new main building in Moelv, is still heavily disputed to this day.
This paper intends to use optimization methods with AMPL and the solver CPLEX 22.1.1.0 with the NEOS Server to construct various models: the status quo, the proposed change, an equal opportunity to healthcare, a maximization of population health outcomes by weighting population, and a pseudo facility location model. This is achieved using minimization, set covering, weighted-set covering, and facility location model formulations.
Conclusively, our findings suggest that the decision to remove the Hamar and Gjøvik hospitals for the Moelv alternative will lead to marginally less fair outcomes for some Innlandet residents, with only few municipalities being affected by longer travel times.
Our findings also suggest that while the weighted set covering model best represents the reality of Innlandet hospital distributions, this may have fairness repercussions for those living farther from densely populated cities based on the maximum distance needed to travel to a hospital.
While this thesis is a predominantly exploratory paper, this intends to serve as a reminder to be conscious of the implementation of fairness in both policy decisions and optimization models.There has been significant research related to fairness concerns in the healthcare space, from philosophical questions asking how societies should distribute resources, to papers addressing who healthcare decision-makers should be: the government, healthcare professionals, the public, or other stakeholders. This paper collects theories from distributive justice, procedural justice, and public healthcare decision-making to explore fair, yet feasible, alternatives to best find healthcare solutions for policy makers, in the context of mathematical modeling with optimization.
The Innlandet case study reveals the difficulty of coming to a concrete answer to these questions as the 20-year debate, removing two hospitals in Hamar and Gjøvik municipalities in favor of a new main building in Moelv, is still heavily disputed to this day.
This paper intends to use optimization methods with AMPL and the solver CPLEX 22.1.1.0 with the NEOS Server to construct various models: the status quo, the proposed change, an equal opportunity to healthcare, a maximization of population health outcomes by weighting population, and a pseudo facility location model. This is achieved using minimization, set covering, weighted-set covering, and facility location model formulations.
Conclusively, our findings suggest that the decision to remove the Hamar and Gjøvik hospitals for the Moelv alternative will lead to marginally less fair outcomes for some Innlandet residents, with only few municipalities being affected by longer travel times.
Our findings also suggest that while the weighted set covering model best represents the reality of Innlandet hospital distributions, this may have fairness repercussions for those living farther from densely populated cities based on the maximum distance needed to travel to a hospital.
While this thesis is a predominantly exploratory paper, this intends to serve as a reminder to be conscious of the implementation of fairness in both policy decisions and optimization models
Quantifying the short-term asymmetric effects of renewable energy on the electricity merit-order curve
Get PDFAmidst the growing significance of renewable energy, this paper examines the asymmetric effects of renewable energy on electricity prices and transmission flows in the Nordics using hourly electricity data. Employing a novel panel asymmetric fixed-effects method, we quantify the non-linear impact of renewable generation technologies on the electricity supply curve. Contrary to previous research, our analysis challenges the assumption of wind having symmetric effects in electricity markets. Specifically, we suggest that an increase in renewable energy cannot lead to price reductions of the same magnitude as the price increases caused by a decrease in wind. In addition, we investigate interconnections between regions and explore asymmetries in transmission flows due to wind generation. Our findings reveal the presence of asymmetric effects in the Nordic electricity market, highlighting their significance in achieving a secure electricity system. These results offer valuable insights for governments, policymakers, and market participants for optimizing the electricity generation mix, prioritizing flexible systems, and making informed investment decisions. © 2024 The Author(s)Quantifying the short-term asymmetric effects of renewable energy on the electricity merit-order curvepublishedVersio
Styrets rolle og vurderinger i strategiske fisjonsbeslutninger: En kvalitativ studie
Get PDFHensikten med denne masteroppgaven er å undersøke hvordan et styre arbeider med beslutningen om fisjonering. Vi ser nærmere på hvilke vurderinger som ligger til grunn for beslutningen, samt hvordan styrets beslutningsprosess forløper seg i praksis. Siden det finnes begrenset forskning på styrets håndtering av slike beslutninger, har vi utforsket temaet gjennom en kvalitativ eksplorerende studie basert på intervjuer med åtte nøkkelinformanter. Informantene består av erfarne styremedlemmer og nøkkelpersoner som har vært involvert i syv ulike fisjoner. Dette utgjør datagrunnlaget for studien. Gjennom dybdeintervjuer har vi identifisert flere sentrale aspekter som diskuteres i styrerommet, samt noen karakteristikker ved styrets arbeid som går igjen i flere av de undersøkte fisjonsprosessene.
Vurderingene vi har identifisert som taler for fisjon er i tråd med konsernteori. Disse inkluderer manglende synergier mellom forretningsområdene, fundamentale forskjeller mellom forretningsenhetene, og begrenset verdiskaping fra morselskapet. Vi diskuterer hvordan disse faktorene har vært avgjørende i fisjonene vi har undersøkt og viser hvordan de også gjør seg gjeldende på styrenivå. I tillegg belyser vi andre hensyn som styret må ta for å fatte en fisjonsbeslutning. Dette inkluderer blant annet hvordan initiativet til styret i fisjonsprosessen kan påvirkes av en eierrepresentant. Videre drøfter vi hvorfor en langvarig fisjonsprosess kan være hensiktsmessig. Den omfattende tidsbruken kan delvis forklares med et sterkt fokus på å oppnå konsensus blant styremedlemmene, som er en utbredt praksis i norske styrer. Vi argumenterer for at oppnåelse av konsensus ikke bare er et mål, men også fungerer som et viktig verktøy for å sikre høyere beslutningskvalitet. Spesielt fremhever vi at det er viktig å inkludere tillitsvalgte og ansattrepresentanter i prosessen.
Avslutningsvis peker vår studie på flere interessante områder for videre forskning. Blant annet fremhever vi eierrepresentanten som en sentral initiativtaker i styret, fisjonsprosessen som en langvarig og kompleks prosess, konsensus som både mål og virkemiddel, samt tillitsvalgtes rolle som en sentral aktør i beslutningsprosessen
Deep Learning-Based Stock Price Prediction for Norwegian Companies: A Comparative Studytive Study
No full textThis thesis discusses the applicability of advanced machine learning (specifically GBDT models) and deep learning models in financial forecasting, especially in the setting of stock price prediction on the TITLON dataset. With this representation of the Oslo Stock Exchange data, which has complexities and volatilities intrinsic to financial markets, prediction of the stock prices is quite challenging. That would mean developing and evaluating robust models that are able to capture intricate patterns in data.
The research is done at a holistic level: the performance of many models of machine learning is estimated, including Gradient Boosting Decision Trees like XGBoost and CatBoost, alongside some deep learning state-of-the-art architectures like Self-Normalizing Networks, Grownnet, DCNv2, AutoInt, MLP, ResNet, and FT-Transformer. These models will be contrasted for their capacity to reduce the error in predictions; two metrics we use to assess the performance of models are the root mean squared error and the mean absolute error.
Another important novelty of this research is the application of a feature extraction with CNN for improving the model toward its precision in accuracy. Such results, before and after applying the CNN feature extraction, are rigorously compared to show that most of the models—deep learning architectures like SNN, DCNv2, AutoInt, and MLP—are substantially improved by this method. These models depicted a large reduction in RMSE and MAE, which goes in line with improved predictive accuracy and generalization ability to unseen data (the test set).
Models such as ResNet and FT-Transformer demonstrated smaller improvements after feature extraction, proving that model-specific tuning could be required to some extent. GBDT models had massive performance improvements, particularly on the MAE, hence proving their robustness and competitiveness toward financial forecasting tasks when matched with efficient features.
This thesis therefore concludes that although CNN-based feature extraction has the capacity to increase a model's predictive performance significantly, the choice of the appropriate model architecture and its hyperparameter tuning remain a prerequisite for optimal results. The findings enrich the literature on deep learning and machine learning techniques applied to financial forecasting, including valuable insights for future studies and practical implementations within the financial industry.This thesis discusses the applicability of advanced machine learning (specifically GBDT models) and deep learning models in financial forecasting, especially in the setting of stock price prediction on the TITLON dataset. With this representation of the Oslo Stock Exchange data, which has complexities and volatilities intrinsic to financial markets, prediction of the stock prices is quite challenging. That would mean developing and evaluating robust models that are able to capture intricate patterns in data.
The research is done at a holistic level: the performance of many models of machine learning is estimated, including Gradient Boosting Decision Trees like XGBoost and CatBoost, alongside some deep learning state-of-the-art architectures like Self-Normalizing Networks, Grownnet, DCNv2, AutoInt, MLP, ResNet, and FT-Transformer. These models will be contrasted for their capacity to reduce the error in predictions; two metrics we use to assess the performance of models are the root mean squared error and the mean absolute error.
Another important novelty of this research is the application of a feature extraction with CNN for improving the model toward its precision in accuracy. Such results, before and after applying the CNN feature extraction, are rigorously compared to show that most of the models—deep learning architectures like SNN, DCNv2, AutoInt, and MLP—are substantially improved by this method. These models depicted a large reduction in RMSE and MAE, which goes in line with improved predictive accuracy and generalization ability to unseen data (the test set).
Models such as ResNet and FT-Transformer demonstrated smaller improvements after feature extraction, proving that model-specific tuning could be required to some extent. GBDT models had massive performance improvements, particularly on the MAE, hence proving their robustness and competitiveness toward financial forecasting tasks when matched with efficient features.
This thesis therefore concludes that although CNN-based feature extraction has the capacity to increase a model's predictive performance significantly, the choice of the appropriate model architecture and its hyperparameter tuning remain a prerequisite for optimal results. The findings enrich the literature on deep learning and machine learning techniques applied to financial forecasting, including valuable insights for future studies and practical implementations within the financial industry
Implementering av kunstig intelligens i offentlig sektor: Endringsledelsens rolle for suksess
Get PDFFormålet med denne masterutredningen er å utforske hvordan endringsledelse kan brukes som et verktøy for å håndtere utfordringer og fremme en vellykket implementering av kunstig intelligens (KI) i offentlig sektor. Fremveksten av KI gir store muligheter for verdiskaping og bedre kvalitet på tjenestene som leveres, men stiller samtidig krav til ledelse for å håndtere komplekse utfordringer. Med riktige verktøy fra endringsledelse kan ledere forebygge motstand, bygge nødvendig kompetanse, skape en felles forståelse for den nye retningen, og sikre en smidig overgang til en KI-drevet arbeidshverdag.
Studien fokuserer på organisatoriske utfordringer som motstand mot endring, mangel på kompetanse og behovet for organisatorisk tilpasning. Kotters åtte-stegs modell og ADKAR-modellen brukes som teoretiske rammeverk for å analysere hvordan ledere navigerer gjennom disse utfordringene. Studien diskuterer også potensialet for smidig endringsledelse som et supplement til tradisjonelle modeller.
For å undersøke dette, ble det gjennomført en kvalitativ casestudie av fem norske offentlige organisasjoner, hvor vi utforsket hvilke utfordringer de har møtt i deres arbeid med implementeringen av KI, og hvordan de har håndtert disse ved hjelp av endringsledelse. Dataene ble samlet inn gjennom intervjuer med ledere og nøkkelpersoner som har vært involvert i KI-implementeringen, samt interne dokumenter. Funnene indikerer at effektiv kommunikasjon, tidlig involvering av ansatte, bruk av tverrfaglige team og endringsagenter, synliggjøring av små gevinster, kulturbygging, samt kompetanseheving er avgjørende for å overkomme barrierer og fremme en vellykket implementering. Funnene viser også variasjon i hvordan organisasjonene tilpasset seg, avhengig av ressursgrunnlag og ledelsens engasjement.
Basert på studiens innsikter anbefales det at organisasjoner utvikler en helhetlig strategi som balanserer perspektivene knyttet til , da suksessen med implementeringen av KI i stor grad avhenger av organisatoriske tilpasninger og aktiv deltakelse fra ansatte. For å møte disse utfordringene bør ledelsen hente ut de mest relevante elementene fra endringsmodellene og tilpasse dem til organisasjonens behov for å sikre en mer helthetlig tilnærming
Får investorer den impacten de forventer fra artikkel 9-fond?
Get PDFBærekraftige investeringer er en investeringsform man forventer vil bidra til å gjøre verden et bedre sted – investeringene skaper positiv impact. I Norge tilbys det fond som utelukkende består av bærekraftige investeringer, og som dermed rapporterer etter artikkel 9 i SFDR. I finansmarkedet markeres fondene ofte med en mørkegrønn farge, og dermed trekker mange parallellen mot bærekraftige resultater. Ved å intervjue forvaltere og ESG-spesialister, samt analysere tilhørende fondsinformasjonkartlegger vi hvorvidt fond som rapporterer etter artikkel 9 i det norske markedet innehar karakteristikaene for impact-investeringer basert på bestemte kriterier. Vi undersøker hvorvidt fondene fører til slike resultater gjennom deres praksis rundt screening, aktivt eierskap og exit. Våre resultater indikerer at kun 16% av fondene skaper impact gjennom sine investeringer. Impact-fondene er utelukkende PE- og VC-fond, og selv om verdipapirfond kan ha påvirkning sammen med andre, virker de ikke å ha samme mulighet til å oppfylle karakteristikaene. Med utgangspunkt i at investorer forventer at fond som driver med bærekraftige investeringer skaper impact, konkluderer vi med at investorer generelt sett ikke får den impacten de forventer fra investering i slike fond
Financial Market Reactions to the 2022 Inflation Reduction Act: A Global Perspective
Get PDFThe Inflation Reduction Act (IRA) of 2022 is the most ambitious climate policy in U.S. history. The law introduces significant subsidies for green companies to combat climate change while also focusing on securing domestic production and reducing reliance on foreign supply chains, particularly from China. In this thesis, we conduct an event study to analyze the market reactions in the U.S., Europe, and Asia following the announcement of the IRA.
Our findings show that U.S. industries focusing on renewable technologies experienced significant positive abnormal returns, reflecting optimism surrounding the IRA’s provisions. Despite hard criticism from EU and Asian leaders accusing the IRA of violating WTO trade rules, our results reveal a more nuanced picture. In Europe, we observe positive market reactions in several renewable-focused industries, suggesting optimism linked to increased global demand for renewable technologies and perceived strengthened competitiveness relative to Asia. In contrast, we observe limited reactions in Asian industries, except in the EV industry, which saw a significant negative reaction.
Our findings provide valuable insights into how financial markets adapt to climate policies and protectionist regulations, offering a valuable perspective for shaping future climate regulations and international trade agreements
The Role of Currencies in M&A
Get PDFThis thesis examines how currency fluctuations affect cross-border mergers and acquisitions (M&A), focusing on the differences between emerging and developed economies. We analyze a dataset comprising 124,381 M&A deals, which includes 23,458 cross-border transactions, across 36 countries from 1999 to the second quarter of 2024.
We study how currency appreciation, depreciation, and volatility influence international investment decisions. We find that currency appreciation in the target country significantly increases cross-border M&A activity. This suggests that a stronger currency indicates robust economic conditions, making the country more attractive to foreign investors. In contrast, higher currency volatility negatively impacts cross-border M&A, deterring investments, as it can be perceived as an increase in risk and uncertainty.
We demonstrate that emerging markets are more sensitive to currency movements than developed markets. We argue that this sensitivity is due to the greater instability in emerging economies
Private Equity, Public Wealth: Assessing the Non-Financial Risks of NBIM’s Potential PE Entry
Get PDFIn the spring of 2024, the Norwegian GPFG (Oil Fund) was once again denied permission to invest in Private Equity (PE). However, the decision is not final, with an Expert Council being appointed to evaluate new investment opportunities for the Fund, including PE.
The purpose of this thesis is to examine the non-financial risks associated with PE and determine whether they make PE an unsuitable investment option for the GPFG. We begin by providing a detailed overview of the GPFG and the PE industry, emphasizing the unique challenges that non-financial risks pose in this context. The goal is to provide the reader with an overview of the arguments in support of and against the decision. Using insights from existing research, a public survey and expert interviews, we will attempt to answer the research questions and conclude whether the non-financial risks of PE complicate an entry for the Fund, and whether they can be mitigated.
There have been limited attempts at categorizing non-financial risks. This thesis therefore introduces a new framework dividing them into five categories. Findings from our research are grouped and analyzed according to this framework.
Arguments in support of a PE entry include a growing market share, the possibility of obtaining greater returns than within public equities, and diversification. On the other hand, critics point to high and sometimes unpredictable costs, greater opacity, and risks to the Fund’s reputation and legitimacy. There is also disagreement regarding the existence of abnormal returns. Finally, some experts believe PE to be a viable investment opportunity, but a bad fit for the GPFG due to its structure and unique characteristics.
We find that the Norwegian general public is skeptical to the ideas of reduced transparency and high fees paid to fund managers. Experts are divided regarding the importance of various non-financial risk factors and whether they can be mitigated. Ultimately, we conclude that NBIM possesses the necessary expertise and governance mechanisms to address these risks. We therefore recommend a PE entry if public support can be secured and if it offers greater financial benefits than other alternatives, such as increasing the equity portion
Fra prosjekt til produkt: innvirkning på jobbtilfredshet
Get PDFDenne masterutredningen utforsker hvordan overgangen fra prosjekt- til produktledelse påvirker de ansattes jobbtilfredshet. Produktledelse, som innebærer en inkrementell og tverrfaglig tilnærming til produktutvikling, er et forskningsområde som har fått svært begrenset oppmerksomhet. Samtidig vil studien bidra med forskning innen jobbtilfredshet, som på den annen side har mottatt betydelig oppmerksomhet i forskning. Denne studien bidrar med empiri til feltet gjennom åtte semistrukturerte intervju med ansatte i en bedrift. Funnene viser at produktledelse har både positive og negative implikasjoner for de ansattes jobbtilfredshet, men hovedsakelig positive.
Forskningen vår viser til at overgangen til produktledelse kan øke jobbtilfredsheten særlig gjennom produktteamene og deres tette samarbeid. Det mener vi kan forklares gjennom styrket samholdet, tilhørighet og tillit. Videre vil det at teamene er tverrfaglige og varige, øke utviklingsmulighetene for de ansatte særlig ved at de tilegner seg etterspurt kompetanse. De ansatte trekker videre frem at å utvikle produktene basert på direkte tilbakemeldinger fra kundene, oppleves svært motiverende. Tilbakemeldingene opplever de som meningsfulle og engasjerende, fordi det bidrar til at produktene samsvarer med kundenes faktiske behov, og styrker følelsen av at arbeidet er viktig gjennom anerkjennelse fra kundene.
Samtidig avdekker vi sentrale utfordringer som knytter seg til antall mål, antall avgjørelser som må fattes, kommunikasjon og tidspress. I tillegg oppleves ledelsen, til tider, som detaljstyrende ved at de pålegger teamene mål. Dette har ført til svekket autonomi og frustrasjon, særlig når målene oppleves som urealistiske. På den annen side kan bruk av OKR og produktstrategier bidra til økt autonomi og tydeligere målsetninger. Informantene har opplevd utfordringer knyttet til kommunikasjon og koordinering, grunnet uklare roller og ansvarsområder. Imidlertid har implementeringen av nye digitale verktøy hjulpet med å strukturere arbeidet og styrke samarbeidet. I tillegg har det de ansatte opplevd det utfordrende å skulle fatte et betydelig antall avgjørelser. Videre har produktledelse medført økt tidspress, men ledelsen har her vært lydhøre og utvidet tidsfristene for produktene, som minket presset. Likevel innebærer tilnærmingen tydelige frister, som kan oppleves som tidspress for enkelte.
Samlet sett peker vår forskning mot at produktledelse kan øke jobbtilfredsheten blant de ansatte. Samtidig krever det at organisasjonene utvikler klare prosedyrer, retningslinjer, roller og ledelsespraksiser for å kunne gi ønsket effekt og resultat