Skip to main content
Login | Suomeksi | På svenska | In English

Browsing by Title

Sort by: Order: Results:

  • Mäkinen, Pyry (2023)
    The two-dimensional path-planning problem involves determining the shortest path between two points on a plane while adhering to a set of polygonal constraints. In this thesis, we show how the path-planning problem in dynamically updated environments gives rise to a need for an edge-flipping based line-segment insertion procedure for constrained Delaunay triangulations, and present an efficient algorithm for the problem. The algorithm operates on convex vertices situated along the boundary of the channel of triangles that intersect with the edge that is to be inserted. We show that such vertices can be removed from the channel using a finite series of edge-flip operations, and that the repeated application of the vertex removal procedure results in the insertion of the requested edge in the triangulation. Furthermore, we present a working sample implementation of the resulting algorithm.
  • Häkkinen, Sanni (2024)
    Cities are culmination points for linguistic diversities, and the variety of spoken languages is encountered in the everyday urban streetscapes. The coming together of cultures in this sense is an outstanding example of Doreen Massey’s throwntogetherness, the question of our living together. Hearing different languages can be a significant part of the sensory experience of a space, and at the same time a means of making sense of the entire social environment. Linguistic landscapes, i.e. language displays in spaces, have been studied almost predominantly by sociolinguists with a focus on the textual domains, whereas the study of the oral domain has remained limited. In more recent research, the subfield of linguistic soundscapes has aimed at better addressing the rich linguistic varieties in everyday contexts. These linguistic soundscapes are hyper-relational and thus ever-changing, constantly negotiating presences and absences. This thesis focuses on studying the different manifestations that linguistic diversities can take in the soundscape, and at the same time, how these diversities can be captured. Through a case analysis of three linguistically diverse areas in Helsinki, I explore and construct an example of conducting linguistic soundscaping. Following the conceptualisation of linguistic soundscaping, suggested by Scarvaglieri et al. (2013), the analysis builds on both qualitative and quantitative methods. I developed an observation method for capturing diversity on a spatially accurate scale by walking in the streetscape and documenting encountered languages. In doing so, I acted as a mediator in the data collection and produced a quantitative snapshot of possible linguistic diversities one could encounter while walking in the study areas. In addition, I utilised interview transcriptions and thematic analysis to understand how linguistic encounters are perceived. The results suggest that the linguistically diverse case study areas (Aleksanterinkatu, Kauppakartanonkatu and Pihlajamäki) show diversity in the spoken languages, and that the narrated linguistic experiences can come from a complex register of ‘knowing’ that builds on familiarity, linguistic resources and past experiences. The observational method, although considered to be challenging concerning subjectivity, soundscape elements and mobility, could be repeated with the findings of this study especially with additional resources. On the other hand, the differing experienced diversities invite further qualitative research, considering the multiple effects of the immediate surroundings in more detail. At best, systematically documented linguistic soundscapes could support understanding of the overall social diversity beyond register data and provide clues of who are audibly present – and absent – in the public spaces.
  • Koho, Mikko (2015)
    Biologisten havaintoaineistojen julkaiseminen linkitettynä datana mahdollistaa useiden aineistojen yhdistämisen toisiinsa. Yhdistämällä toisiinsa useita samaan asiaan liittyviä aineistoja, voidaan saavuttaa parempi ymmärrys kiinnostuksen kohteena olevasta ilmiöstä kuin tutkimalla aineistoja erikseen. Näin voidaan mahdollistaa tarkempien päätelmien tekeminen aineistojen pohjalta sekä etsiä odotettuja tai odottamattomia yhteyksiä aineistojen välillä. Linkitetyssä datassa käytetty RDF-tietomalli tuo aineistoihin koneluettavuuden ja helpon tavan viitata kaikkiin aineistojen osiin. Linkitettynä datana julkaistuja aineistoja voidaan helposti rikastaa yhä uusilla aineistoilla. Tässä tutkielmassa käsitellään Hangon lintuaseman havaintoaineiston sekä Ilmatieteenlaitoksen Hangon Russarön säähavaintoaineiston mallinnusta, käsittelyä ja hyödyntämistä linkitettynä datana. Aineistot on mallinnettu käyttäen RDF Data Cube -sanastoa, joka parantaa aineistojen yhteentoimivuutta. Lintuhavaintoaineistoon on annotoitu lajitietoa käyttäen ontologiaa Suomen linnuista, jota on rikastettu mm. lajien tuntomerkkiontologialla sekä uhanalaisuustiedoilla. Aineistot on julkaistu Linked Data Finland -alustalla, ja aineistojen välisten yhteyksien hahmottamiseksi on kehitetty visualisointipalvelun prototyyppi. Säätilan tiedetään olevan tärkeimpiä päivittäisen lintumuuton voimakkuuteen vaikuttavia tekijöitä. Visualisointipalvelulla pyritään näyttämään käyttäjälle, miten säätila vaikuttaa lintuhavaintomääriin ja erityisesti havaittuun lintumuuttoon. Aineistojen välisten suhteiden parempi tuntemus mahdollistaa tarkempien päätelmien tekemisen lintuhavaintoaineiston perusteella. Tutkielmassa esitetyt menetelmät ovat yleistettävissä lintu- ja säähavaintoaineistojen lisäksi muihin rakenteeltaan samankaltaisiin aineistoihin.
  • Somero, Sonja (2018)
    Sovelluksen ylläpitäminen on usein kallista. Muutosten tekeminen ja vikojen korjaaminen on edullisempaa kehitysvaiheessa kuin ylläpitovaiheessa. Jotta mahdollisimman moni virhe löytyisi aikaisessa vaiheessa, ohjelmistoprojekteissa voidaan käyttää muun muassa staattisen analyysin työkaluja, joilla virheitä voidaan havaita. JavaScript on yksi käytetyimmistä ohjelmointikielistä, mutta se sisältää kuitenkin joukon huonoja käytänteitä. Lisäksi jokin asia, kuten funktion määrittely, voidaan toteuttaa usealla eri tavalla. Huonojen käytänteiden välttämiseksi ja ohjelmakoodin yhdenmukaisuuden lisäämiseksi JavaScriptille on kehitetty lint-työkaluja, jotka tekevät ohjelmakoodille staattista analyysia etsien edellä mainittujen ongelmien lisäksi syntaksivirheitä ja muita mahdollisia virheitä. Tässä tutkielmassa tarkastellaan miten lint-tyyppisen ESLint-työkalun käyttöönottaminen vaikuttaa JavaScript-sovelluksen ylläpidettävyyteen. Tutkimus tehdään yksittäiselle projektille, jossa toteutettiin JavaScript-sovellus. Ylläpidettävyyttä verrataan metriikoiden avulla ennen ESLintin havaitseminen ongelmien korjaamista sekä korjausten jälkeen. Lisäksi kohdeprojektin kehittäjä arvioi ESLintin avulla korjattua ohjelmakoodia. Tulosten perusteella ESLintin käyttöönotto on helppoa ja nopeaa ja sen käyttämisen katsotaan parantavan ylläpidettävyyttä. Lisäksi ESLintin käyttämisestä seurasi, että aiemmin piilossa olleet toisteiset rakenteet tulivat ohjelmakoodin yhdenmukaistumisen myötä esille.
  • Liljestrand, Hans (2017)
    Linux kernel vulnerabilities are often long lived and in some cases challenging to patch after discovery. The current focus in upstream Linux security has therefore been on categorical protections against whole error classes, not only reactive patching of specific vulnerabilities. Our work contributes to these efforts by tackling memory errors in the Linux kernel from two different fronts. First, we contributed to the upstream Linux kernel by working on a mechanism to prevent use-after-free errors caused by reference counter overflows. Second, we explored the applicability of Intel MPX as a general mechanism to prevent spatial memory errors in the Linux kernel.
  • Lishchuk, Valeriia (2024)
    In the theoretical part of the thesis, the main trends in the study of lipid aggregates as a pseudostationary phase in capillary electrokinetic chromatography (EKC) over the past 15 years is reviewed. Due to ability of liposomes to mimic cell membranes, they have found wide application in the biomedical field and analytical chemistry. Currently, various liposomal formulations are used as background electrolyte solutions in liposome EKC. The possibility to vary the lipid content in the lipid aggregates is increasingly expanding the usage of liposome for solving specific tasks. One of the main components of vertebrate cell membranes is cholesterol. The enormous effect of this sterol on the bilipid layer has been well studied. Properties of the lipid layer in the liposomes can be altered by changing the molar percentage of cholesterol and by this it is possible to regulate and bring the experimental conditions closer to biological conditions. In the kingdom of fungi, the function of cholesterol is performed by ergosterol. The similarity of the sterols suggests that compounds would have a similar effect on the lipid bilayer. However, comparative analyses of systems involving cholesterol and ergosterol show less obvious results. Understanding the mechanisms of drug interactions with ergosterol-containing liposomes may provide ideas for the potential development of e.g., antifungal drugs. The experimental part of the thesis describes LEKC experiments with five steroids, namely 21-deoxycortisol, corticosterone, testosterone, 17α-hydroxyprogesterone, and progesterone. Two different sets of liposome formulations containing ergosterol or cholesterol were investigated. A total of nine pairs of both liposome dispersions were used as background electrolytes for the study the interactions between the liposomes and the steroids. The experimentally determined retention coefficients and the corresponding calculated distribution constants (KD) reflected the dependency of the interaction with the lipophilicity of the analytes. With increasing lipophilicity (hydrophobicity), the affinity of the steroids to the liposomes increased, as seen by stronger interactions with the lipid phase. The average relative standard deviation of the distribution constants for each steroid was not more than 5.5% for the cholesterol-rich liposomes and 4.3% for the ergosterol-rich liposomes. The regression model of log KD Ergo vs. Log KD Chol demonstrated a linear correlation between the obtained data. The linear determination coefficient (R²) was 0.9254. Pearson's linear correlation coefficient reflected also a strong linear correlation (r = 0.9615).
  • Eurasto, Felix (2024)
    G protein-coupled receptors (GPCRs) constitute the largest family of receptors in humans. They are involved in the regulation of major biological processes including sight, taste, and mood. Due to their prevalence in the human body and involvement in such a wide range of tasks, GPCRs are medically extremely important. GPCRs are cell-surface receptors, responsible for conveying biological messages from the extracel- lular domain to the cytoplasmic region. As such, GPCRs are constantly interacting with the lipids of the cell membrane. These interactions are thought to mediate the activation behaviour of the GPCRs, although the exact nature of these effects is often unknown. The beta-2 adrenergic receptor (β2AR) is a class-A GPCR, whose native ligand is adrenaline. It plays a crucial role in the inactivation of the sympathetic nervous system to trigger the fight-or-flight response. Many GPCRs exhibit basal activity. That is, these receptors can activate even in the absence of an activating ligand. β2AR is one of these GPCRs. The specific cause and mechanism of basal activity are often unknown and, as of the start of the project presented in this thesis, were undetermined for β2AR. We used high-throughput fully atomistic molecular dynamics (MD) simulations coupled with ma- chine learning (ML) methods to ascertain specific interactions between a highly conserved aspartate residue of β2AR and phosphatidylcholine (PC) lipids that stabilize the active state of β2AR. We also found that cholesterol plays a role in mediating these interactions. These results shed light on the effect of the lipid composition of one’s cell membranes, and by extension one’s lipid diet, on the activation behaviour of β2AR, a medically extremely relevant receptor.
  • Pöhö, Päivi (2013)
    Lipidomics or global lipid profiling is a branch of metabolomics that aims to comprehensive analysis of lipids in a biological system. Lipidomics has become an important research field, since the increased awareness of lipid functions in cell and their role in many common diseases. Also the development of analytical methods, especially mass spectrometry has driven the progress of lipid profiling methods. Lipid profiling allows detection and quantitation of hundreds of intact lipid species in parallel. Challenges in lipidomics are the diversity of different lipid structures and varying lipid concentrations in biological samples. Despite of the development of lipidomics there is not yet a single analytical method to screen all lipids in a biological system. Therefore targeted lipidomics methods are needed in addition to global lipidomics. The literature part of this study presents different analytical methods used in lipidomics studies. The most popular lipid profiling methods are liquid chromatography-mass spectrometry (LC-MS) and direct infusion mass spectrometry or so called shotgun analysis. LC-MS based methods usually utilize reversed phase columns including octadecylsilane stationary phases, but also normal phase chromatography has been applied. In shotgun lipidomics crude extracts are infused directly to mass spectrometer without prior separation. In lipidomics, both unit resolution mass spectrometers like triple quadrupole and ion trap as well as high mass accuracy mass spectrometers like Fourier transform ion cyclotron resonance, Orbitrap, and time of flight mass spectrometers have been used. Mass scanning methods have provided full scan acquisition with accurate mass as well as tandem mass spectrometry like precursor ion scan, neutral loss scan, and multiple reaction monitoring. From the atmospheric pressure ionization techniques, electrospray ionization has been the most applied ionization, but also other soft ionization techniques such as matrix-assisted laser desorption/ionization has been employed. Sample pretreatment in global lipid profiling relies on homogenization and extraction with chloroform-methanol. The most used extraction methods are Folch and Blight and Dyer extractions or modifications of these. In addition, other sample pretreatment methods like solid phase extraction, derivatization, and thin layer chromatography have been used, especially with targeted lipidomics methods. Lipidomics platforms have been applied to several studies of different diseases like diabetes, schizophrenia, and cancer. Many bioactive signaling lipids cannot be detected with lipid profiling methods, since they exist at low concentration and have polar structure. In the experimental part a targeted method was developed for analysis of signaling lipids by ultrahigh-performance liquid chromatography-mass spectrometry. Selected compounds were lysophosphatidic acid, sphingosine-1-phosphate, and phosphoinositides. Method development was mainly done with triple quadrupole mass spectrometer, but also Orbitrap was applied. Several different columns, eluent systems, and pretreatment methods were tested, as well as direct infusion. The best chromatographic separation to signaling lipids was achieved with a reversed phase column at alkaline conditions. However, with this method also several drawbacks were encountered. Peaks were shifting and broadening, there were carryover effects, and problems with repeatability and sensitivity. Direct infusion on the contrary turned out to be problematic because of the unstable electrospray and formation of lysophosphatidic acid from lysophosphocholine in the ionization chamber. Development of a single analytical method for these signaling lipids turned out to be a challenging task and a complete working method for all the studied compounds was not attained.
  • Ravald, Henri (2023)
    In the literature part, different analytical techniques for investigating drug-lipid interactions were reviewed, with the emphasis being on capillary electromigration and liquid chromatog-raphy techniques involving the use of artificial lipid vesicles called liposomes. Special interest was reserved for studies on beta blockers, a class of cardiovascular drugs. In addition, com-mercial intravenous lipid emulsions were introduced and in vivo and in vitro research on them was reviewed, mainly focusing on studies of their application in treatment of drug over-doses. Other methods for studying drug-lipid interactions were also briefly discussed and compared to electrophoretic and chromatographic techniques. In the experimental part, the interactions of seven beta blockers with a commercial intrave-nous lipid emulsion, Intralipid, were studied by a capillary electrophoresis technique called liposome electrokinetic chromatography. Electrophoretic mobilities of the analytes in a capil-lary filled with Intralipid solution were determined. From the mobilities of the analytes and Intralipid solution, retention factors and distribution coefficients for the beta blockers were determined. The values were compared with literature octanol/water partition coefficients and experimentally determined partition coefficients from previous research. The particle sizes of the lipid emulsion were studied by dynamic light scattering and flow field-flow fractionation. Other capillary electromigration techniques such as partial-filling electrokinetic chromatog-raphy and capillary electrophoresis frontal analysis were also briefly performed, and the results were compared with liposome electrokinetic chromatography. Also, interactions of the In-tralipid solution with the silica capillary wall were studied.
  • Ghasemi-Vierling, Nasim (2017)
    Condensed water molecules need a surface such as liquid or solid aerosol particle surfaces to accommodate. One factor that determines the growth of submicrometer aerosol particles during the process of water vapor molecule absorption is the mass accommodation coe cient. In this thesis previous works on computer modeling and measurement of mass accommodation are reviewed.
  • Åberg, Susanne (2013)
    The Litorina Sea is the latest Baltic Sea stage, which has dated to begin about 9000−7500 years ago. The sea phase developed after the Ancylus Lake when the straits of Denmark opened, and it is a distinctive phase within the Baltic Sea history. This MSc thesis aims to create a GIS based database about the highest stand of the Litorina Sea in Finland and to reconstruct its extent by interpolating it's shorelines with the resolutions of 10 and 25 meters. The purpose was also to make an updated version of the highest Litorina Sea shoreline map by Eronen and Haila 1992. The highest shoreline of the Litorina Sea can be observed in ancient shoreline relicts such as stone belts, beach ridges, cliffs and terraces. The widest extent of the Litorina Sea's brackish water can be detected from isolation horizons within mire and lake sediments. It can be distinguished as Clypeus limit or physically as a chance where the clays of the Ancylus Lake turn into clays of the Litorina Sea. The highest shoreline of the Litorina Sea is diachronous and its level in different regions of Finland varies due to the uneven uplift and inclining of the shorelines. It is important to note that the highest shoreline of the Litorina Sea has occurred in some regions at the beginning of the Litorina Sea stage like in Ostrobothnia or later by the time of the Litorina transgression in other places like in southern Finland. Depending on the location, the highest shoreline of the Litorina Sea occurred about 8000−6500 year ago due to its diachronity and the impact of transgression. All known shoreline and isolation observations were collected from the literature in the database with the help of LiDAR DEM with a resolution of two meters, and DEM 25. Observations were used to create a database with the information of coordinates and observed altitude. The aim was also to find new shoreline sites. Altogether 564 shoreline observations and 102 isolation observations were collected in the database and they were classified to five classes according to their reliability. The highest shoreline of the Litorina Sea (LI) was reconstructed with the help of the best observation points by using vector based TIN interpolation. Two different resolution LI representing rasters were created by subtracting the formed TIN model from two separate DEM models: DEM 10 and DEM 25. Calculated raster grids represent the relative altitude and the depth of the Litorina Sea at its highest stage in different areas. These rasters were classified into two classes, which other class represents area covered with water and other represents land area, and they were converted into polygons. Modeling was challenging because observations were not evenly distributed among the sphere of influence of the Litorina Sea. The altitude of LI increases in north-west direction from about 21 m in Virolahti area to about 32 m in Helsinki area, 45-52 m in Turku and Pori and to 100 m in Ostrobothnia area. Highest observation altitude of the Litorina Sea is in Ylitornio in the slopes of the Huitaperi at the level of 108.4 m. The results of the modeling resemble earlier maps of the highest shoreline of the Litorina Sea but they are more detailed. On the other hand the 25 m resolution map is still very coarse reflecting the errors of DEM 25 model. However, the 10 m resolution map is rather detailed, but it still needs some enhancement. Ancient Shoreline Database and associated fine-resolution maps are published to facilitate and promote the research related to the highest stage of the Litorina Sea.
  • Autio, Mika (2017)
    Liukuvalu on betonirakentamisessa tekniikka, jossa betoni valetaan muottiin kerroksittain siten, että pohjimmaisten kerroksien kovettuessa niiden pinnalle voidaan vielä lisätä betonia ilman saumojen muodostumista. Näin muottia voidaan liu'uttaa ylöspäin uusien kerrosten valmistuessa ja työ etenee yhtäjaksoisesti jo kovettuneen betonin kannatellessa muottia ja ylle valettavia betonikerroksia. Liukuvalua käytetään usein tornien ja massiivisten siilojen rakentamisessa ja siksi siinä onnistuminen on erityisen tärkeää. Liukuvalussa muottia on siirrettävä sopivalla nopeudella, että alemmat kerrokset eivät romahda, mutta valupinta ei pääse myöskään tarttumaan muottiin ja aiheuttamaan repeämiä. Tutkimuksen tilasi ja kustansi Rudus Oy. Tutkimuksen tavoitteena oli selvittää liukuvalubetonoinnin onnistumiseen vaikuttavia tekijöitä. Tärkeimpinä tarkastelukohteina olivat vesi-sementtisuhde, eri sementtilaatujen toimivuus, yleisimmät lisäaineet kuten notkistin ja huokoistin, sekä pozzolaanisten seosaineiden vaikutus. Tutkittavia seosaineita olivat masuunikuona, lentotuhka ja piidioksidijauhe. Kokeilla haluttiin mallintaa todellista rakennustilannetta ja saadun tiedon oli oltava käytännön tilanteissa sovellettavaa. Tutkimus tehtiin Rudus Oy:n tutkimuslaboratoriossa Helsingissä. Tärkeimpänä aineistona tutkimuksessa käytettiin Kjell Fåre Tossan väitöskirjaa Slipforming of Vertical Concrete Structures. Friction between Concrete and Slipform Panel (2001). Kjell Fåre Tossa esitti väitöskirjassaan liukuvalubetonoinnin onnistumisen tärkeimmäksi tekijäksi liukuvalumuotin ja betonin välisen kitkan. Tämän ja muun kerätyn aineiston perusteella rakennettiin mittauslaitteisto, jolla liukuvalua tutkittiin. Laitteistolla mitattiin betonin ja liukuvalumuotin välille kehittyvää kitkaa, seurattiin betonin kovettumista sekä tarkasteltiin valetun pinnan siisteyttä. Betonin kovettumista tarkkailtiin penetrometrillä, lämpötila-anturilla ja valun yhteydessä valmistetuilla koekappaleilla. Betonin ja liukuvalumuotin välistä siirtokitkaa verrattiin betonin kovettumisen eri vaiheisiin ja betonin eri komponentteja säädeltiin niiden annostelun vaikutuksen selvittämiseksi. Kokeiden perusteella pääteltiin, että suurin ongelmia aiheuttanut tekijä betonissa oli huokoistimen annostelu. Huokoistuksella betoniin tehdään pieniä ilmahuokosia, jotka parantavat sen pakkasenkestävyyttä. Käytetty huokoistin oli rasvahapposaippuapohjainen vesiliuos, joka heikensi muottiöljyn toimivuutta ja aiheutti betonisaarekkeiden kasvamista liukuvalumuottiin. Tämä aiheutti betonin repeytymistä liukuvalumuottia siirrettäessä. Muista tekijöistä tärkeimpänä oli reseptin vesi-sementtisuhde. Pienemmällä vesi-sementtisuhteella valmistettu betoni tuotti enemmän kitkaa muottia siirrettäessä ja lisäsi repeämien syntymisen riskiä. Seosaineista piidioksidin havaittiin kasvattavan liukukitkaa ja masuunikuonan pienentävän sitä.
  • Viertola, Julia (2022)
    Pastoral livelihoods are currently changing in Kenya. Economic, societal and environmental development of the country has led to increased standard of living and demand of livestock products in the markets. The rising demand to contribute to food production puts pressure on pastoralist production, while keeping livestock is becoming more challenging due to changes in land use, commercialisation and climate change. Pastoralists are at crossroads, in which they should adapt to the mentioned changes through options such as sedentarisation and livelihood diversification while taking gender into account. Pastoralists are relatively vulnerable due to being a socially marginalized group in Kenya and several development actors are now focusing their projects on pastoralists in order to support them in adaptation. However, Mausch et al. (2021) have argued that development projects often neglect the needs of the target group. Including the target group’s opinions and aspirations to project objectives can lead to more inclusive and sustainable outcomes. ESSA - Earth observation and environmental sensing for climate-smart sustainable agropastoral ecosystem transformation in East Africa is an example of a development project which could benefit from acknowledging target group’s aspirations. Within the systemic change of pastoralism it is crucial to look more deeply at separate pastoral communities. Pastoralism is connected to environment, culture and history which vary between locations, thus they cannot be categorised as a homogeneous group. However, most pastoral communities share the traditional labour division between women and men. In patriarchal pastoral communities women have traditionally been responsible for work labelled as reproduction. Pastoral women have not had the same opportunities in life or representation in society as pastoral men or women in urban areas which puts them in a doubly marginalised position. In this study I combine mapping livelihood aspirations of women and exploring womanhood in the context of pastoralism. The choice to focus on pastoral women’s livelihood aspirations is influenced by the field of feminist geography, the lack of information on pastoral womanhood, the ESSA project and my personal interest towards strengthening the position of women in Global South. The aim of this research is to explore livelihood aspirations of six pastoral Maasai women in Taveta Sub-County, Kenya. Additionally, I am searching for possible connections between the aspirations and the women’s current positions and roles. This research also contributes to the ESSA project by providing information on pastoral womanhood and women’s opinions towards beekeeping, because ESSA’s women-related objectives include livelihood diversification through beekeeping. Feminist geography materialises through aims to broaden the understanding of lives of women, who have traditionally not gained attention in science. This work also includes extensive self-reflection of my position as the researcher, which is instrumental in feminist geography. The research material was collected through fieldwork which took place in six households close to Salaita Hill and Lake Jipe. The research methods were participative observation and semi-structured interviews which were executed with the support of two translators. The research material consists of field diary, in which I report what I observed, interview transcriptions and ESSA’s project paper. In this research I represented the women’s future livelihood aspirations as the women expressed them, and analysed the type of aspirations. The women wanted to continue pastoralism, although most of them were also interested in diversifying their livelihoods to farming, business and casual labour. The women knew relatively little about beekeeping, and the main message from the interviews was that these Maasai women are afraid of bees. Thus, they were not interested in beekeeping as livelihood. From the field diary and interview transcriptions I interpreted that there were connections between the livelihood aspirations, pastoral culture and its patriarchal characteristics, which determine women’s thoughts. Additionally, it seemed that casual work and lack of education might affect the livelihood aspirations. However, the conclusions of this research are not objective because they are interpretations affected by my subjective positionality as the researcher.
  • Vuori, Arto (Helsingin yliopistoHelsingfors universitetUniversity of Helsinki, 2009)
    Low Level Virtual Machine (LLVM) on moderni koko ohjelman elinkaaren optimointeihin keskittyvä kääntäjäarkkitehtuuri. Java-virtuaalikone on puolestaan suosittu korkean tason virtuaalikone, johon monien ohjelmointikielten toteutus nykyään perustuu. Tutkielmassa esitellään alun perin suorituskykyisen C- ja C++-kääntäjän toteuttamiseksi luotu LLVM-järjestelmä ja arvioidaan, miten hyvin LLVM-infrastruktuuri tukee Java-virtuaalikoneen toteuttamista. Tämän lisäksi tutkielmassa pohditaan, miten dynaamisten kielten usein tarvitsemaa suoritusaikaista ja lähdekieliriippuvaista optimointia voidaan tukea lähdekieliriippumattomassa LLVM-järjestelmässä. Lopuksi tutkielmassa esitellään kehitysehdotelma yleisen roskienkeruuinfrastruktuurin toteuttamiseksi LLVM:ssä, mikä tukisi dynaamista muistia automaattisesti hallitsevien kielten, kuten Javan ja sen virtuaalikoneen toteuttamista.
  • Vuola, Marketta (2015)
    Biodiversity conservation, as human activity, is inherently political. Attempts to preserve species and habitats with strict Protected Areas in the Global South often take place in already inhabited regions. Conservation has been often externally imposed on the local, rural communities, resulting in deprivation of their livelihoods and breaking up of their natural resources management traditions, but also in local opposition that threatens conservation outcomes. However, as rural livelihoods depend directly from their surrounding ecosystems, rural communities and conservation planners are argued to have substantial common interests. This study tries to understand the relationship between local communities and conservation authorities in order to create knowledge on how they could form partnerships and work together for natural resources management. This case study took place is Ranomafana National Park in South-Eastern Madagascar where several villages were visited in order to gain knowledge of conservation from the point of view of local communities. The case study follows ethnographic approach using qualitative, semi-structured interviews as the principal method of data collection. Employing a political ecology approach and looking at the institutional arrangements guiding conservation at the local, national and global levels and across formal and informal spheres, this study looks at the power relations in the current forms of co-management and the social impacts they have at the local level. Finally it tries to find out if any form of partnership is formed; if local people are able and willing to manage their natural resources in cooperation with conservation authorities. Although conservation has significant negative impacts on local livelihoods in Ranomafana region, in practice community participation to decision-making is very limited and the local people find it hard to get their voice heard. The results of the case study indicate that the main obstacle for co-management is the failure to respect the rights of local communities to equitable treatment, to recognition as stakeholders, and to participation in decision-making. The feeling of being disrespected creates resentment and mistrust towards conservation authorities. In these circumstances, economic incentives offered do not support community empowerment but rather create dependence from external help. Co-management activities can also enforce the existing inequalities at the local level if only the more powerful segments of communities are included. The case study also shows that local communities are important actors in conservation, able to challenge it – or support it if they view the rules as legitimate – but the actual community self-organising for conservation requires at least some authority over their surrounding ecosystem.
  • Mahmood, Yasir (2018)
    This thesis discusses the notion of locality used in finite model theory to obtain results about the expressive power of first order logic. It turns out that the most commonly used Ehrenfeucht-Fraïssé games are also applicable over finite structures. However, we analyze with an example the need for simpler tools for finite structures due to the complex combinatorial arguments required while using EF-games. We argue that locality is such a tool, although the gap between games and locality is quite narrow as the latter is in fact based on the former. Intuitively speaking: locality of FO implies that in order to check the satisfiability of a FO formula over a finite structure, it is enough to look at a small portion of the universe (which will be called the neighborhood of a point). We discuss two commonly known notions of locality given by William Hanf and Haim Gaifman. We provide the original results of the authors and then their modified versions suitable for finite structures. We then show that first order logic over any relational vocabulary has both of these locality properties. In order to grasp the idea of locality we also include examples wherever required. Towards the end of the thesis we also discuss deficiencies and limitations of the two types of locality and possible solutions to overcome them. In the last section we also discuss locality of order-invariant first order formulas.
  • Viiri, Kalle (2020)
    Locally checkable labeling problems in the LOCAL model of distributed computation are known to have only three distinct complexity classes when the attention is restricted to problems on toroidic grids only: trivial with time complexity Θ(1), local with time complexity Θ(log∗ n) and global with time complexity Θ(n). Prior work shows that problems belonging to the trivial class are easy to recognize, but that local and global labeling problems are undecidable to separate. However, a method called algorithm synthesis exists for creating an asymptotically optimal normal-form algorithm for any locally checkable labeling problem with a time complexity of Θ(log∗ n). This process, when automated, can be used to process vast amounts of suitably encoded labeling problems in bulk, creating a more defined boundary for the undecidable class of local problems. As a proof-of-concept of this method, this work presents a new asymptotically optimal algorithm for a relaxed form of 3-coloring as well as methods for more general search of local problems.
  • Saukkoriipi, Mikko (2022)
    Two factors define the success of a deep neural network (DNN) based application; the training data and the model. Nowadays, many state-of-the-art DNN models are available free of charge, and training and deploying these models is easier than ever before. As a result, anyone can set up a state-of-the-art DNN algorithm within days or even hours. In the past, most of the focus has been given to the model when researchers were building faster and more accurate deep learning architectures. These research groups commonly use large and high-quality datasets in their work, which is not the case when one wants to train a new model for a specific use case. Training a DNN algorithm for a specific task requires collecting a vast amount of unlabelled data and then labeling the training data. To train a high-performance model, the labeled training dataset must be large and diverse to cover all relevant scenarios of the intended use case. This thesis will present an efficient and straightforward active learning method to sample the most informative images to train a powerful anchor-free Intersection over Union (IoU) predicting objector detector. Our method only uses classification confidences and IoU predictions to estimate the image informativeness. By collecting the most informative images, we can cover the whole diversity of the images with fewer human-annotated training images. This will save time and resources, as we avoid labeling images that would not be beneficial.
  • Keller, Levi (2019)
    The spin-orbit-coupled insulator Sr 3 NiIrO 6 is a strongly correlated transition metal compound, where an interplay of geometric frustration and spin anisotropy gives rise to novel magnetic phases. Resonant inelastic x-ray scattering (RIXS) is a powerful probe of the low-lying quasi-particle excitations that underpin these emergent properties. In this work, we partition the active space into approximately non-interacting parts in order to introduce a tight-binding single-particle model Hamiltonian describing the distorted IrO6 octahedra in Sr3NiIrO6. We then use this model to calculate its RIXS spectrum at the Ir L3-edge in the sub-electronvolt range. The results of this calculation are compared with experiments performed at the European Synchrotron Radiation Facility, and with a multiplet crystal field model calculation. We find that this one electron model largely agrees with the full-multiplet model and describes the d-d excitations observed in experiment. The addition of an exchange field term explains the low-lying temperature-dependent magnetic feature, disambiguating the sign of the crystal-field term, and suggesting that the feature is well localized at low temperatures, and is best described as an orbitally- entangled local spin-flip excitation. However, the correspondence at room temperature diminishes, suggesting that dispersive description is necessary to model this regime. The drastic reduction in active space entailed by this model facilitates the creation of extended non-collinear Heisenberg-like models, which can be calculated at a lower computational cost than full multiplet extended models.
  • Koivunen, Vesa (2015)
    Logaritmin käsite on muuttunut sen löytymisen jälkeen huomattavasti. Lähinnä laskennalliseen hyötyyn luodusta apuvälineestä on tullut keskeinen matemaattisen analyysin työkalu. Myös lukuisat sovellukset usealla eri tieteenalalla hyödyntävät logaritmin ominaisuuksia. Tästä syystä se on edelleen oleellinen aihekokonaisuus lukiomatematiikassa. Tutkimuksen tavoitteena on selvittää millä tavoin logaritmi opetetaan koulumatematiikassa. Teoreettisen taustan pohjalta arvioidaan opetuksen nykytilaa ja pohditaan miten sitä voitaisiin kehittää. Tutkielmassa käydään läpi logaritmin opetuksesta tehtyjä tutkimuksia sekä tarkastellaan Helsingin Yliopiston matematiikan aineenopettajaksi opiskelun aloittaneiden ymmärrystä logaritmista. Lisäksi esitellään yhden oppitunnin mittainen opetuskokeilu, jossa hyödynnetään tutkielmassa esiteltyjä keinoja. Tutkimuksen teoreettisena viitekehyksenä toimii David Tallin matematiikan kolme maailmaa. Oleellisessa osassa on proseptin käsite. Proseptuaalisen tiedon tasolla yksilö kykenee ajattelemaan joustavasti proseduraalisen ja konseptuaalisen tiedon välillä. Tallin luoma käsite 'met-before' kuvastaa oppilaan aiemmin kohtaamaa tietoa. Tämä on oleellista opetuksen suunnittelussa, jolloin tulee tarkastella mitkä oppilaan aikaisemmat tiedot toimivat uuden tukena ja mitkä problematisoivat sitä. Logaritmin yhteydessä tarkastelu on erityisen tärkeää, sillä käsitteen historiallinen muodostuminen poikkeaa sen nykyisestä esitysmuodosta. Historiallista kehitystä, oppilaan aiemmin kohtaama tietoa ja logaritmin käsitteen jäsentämistä pyritään helpottamaan ottamalla käyttöön käsitekartta. Tutkimuksessa esitellään logaritmin historiallinen käsitteenmuodostus Antiikin Kreikasta aina John Napierin vuonna 1614 kehittämään käsitteeseen. Tämän lisäksi tarkastellaan logaritmin nykyistä esitysmuotoa ja sen sovelluksia. Logaritmin esiintymistä koulumatematiikassa tutkitaan opetussuunnitelmien ja oppikirjojen perusteella. Perusopetuksen opetussuunnitelman perusteet antavat pohjaa niille tiedoille, jotka lukio-oppilas on aiemmin kohdannut. Lukion opetussuunnitelman perusteita tarkastellaan sen nykyisessä muodossa. Lisäksi huomioidaan tulevan opetussuunnitelman perusteiden luonnos. Tutkimuksessa käydään läpi viimeisen vuosisadan läpi kestänyttä pedagogista keskustelua logaritmien opetuksesta. Tämän lisäksi tehdään katsaus opetuskokeiluihin, joita logaritmin opetuksesta on tehty. Helsingin Yliopiston Matematiikan ja tilastotieteen laitoksen uusilla matematiikan aineenopettajaksi opiskeleville tehdyn kyselyn perusteella arvioidaan yliopistoopinnot aloittavan opiskelijan tasoa logaritmi- ja eksponenttifunktioissa. Tulosten perusteella opiskelijoiden ymmärrys logaritmin suhteen on vahvasti proseduraalisen tiedon tasolla. Sen sijaan eksponenttifuktioissa näkyy konseptuaalisen tiedon tasoa. Lopuksi esitellään yhden oppitunnin mittainen opetuskokeilu logaritmeista. Pyrkimyksenä on esittää kokonaisuus siten, että se on tutkimukseen perehtyneen toteutettavissa. Tässä yhteydessä käydään myös läpi oppitunnista saatua palautetta ja pohditaan muun muassa tutkimuksen linkkejä opetukseen ja mahdollisia jatkotutkimuskohteita.