Traffic speed is generally considered a core issue in roadway safety. Previous studies show that faster travel is not necessarily associated with an increased risk of being involved in a crash. When vehicles travel at the same speed in the same direction (even high speeds, as on interstates), they are not passing one another and cannot collide as long as they maintain the same speed. Conversely, the frequency of crashes increases when vehicles are traveling at different rates of speed. There is no doubt that the greater speed variation is, the greater the number of interactions among vehicles is, resulting in higher crash potential. This research tries to identify all major factors that are associated with speed variation on multilane highways, including roadway access density, which is considered to be the most obvious contributing factor. In addition, other factors are considered for this purpose, such as configuration of speed limits, characteristics of traffic volume, geometrics of roadways, driver behavior, environmental factors, etc. A microscopic traffic simulation method based on TSIS (Traffic Software Integrated System) is used to develop mathematical models to quantify the impacts of all possible factors on speed variation.

En viktig del av den svenska välfärden är kommunens äldreomsorg för personer över 65 år, där Socialtjänstlagen ger rättighet till vårdboende och andra servicetjänster när det behövs. Kommunernas kommunaltjänst ska enligt Socialtjänstlagen vara av god kvalitet. Rapportering av avvikande händelser är en av hörnpelarna i arbetet med att förbättra kvalitet och säkerhet i vården. Avvikelserapportering bidrar till att undvika att negativa händelser upprepas, samt att rutiner förbättras för att höja kvalitet. Genom att ta tillvara möjligheterna med avvikelserapporter kan kvaliteten på vårdboenden förbättras och utvecklas. Denna studie visar på behovet av ytterligare fokus på arbetet med avvikelsehantering gällande omsorg.

Gegenstand der vorliegenden Dissertation ist das Verhalten sogenannter Selbstüberschneidungslokalzeiten $\\|\\ell_t\\|_p^p$ einer zeitstetigen Irrfahrt $(S_r)_r$ auf dem $d$-dimensionalen Gitter $\\Z^d$. Dabei ist für $p>1$ die Funktion $\\ell_t$ definiert durch $$ \\ell_t(z):=\\int_{0}^{t}\\1_{\\{S_r=z\\}}\\,\\d r\\nonumber $$ und bezeichnet die Aufenthaltsdauer der Irrfahrt bis zum Zeitpunkt $t\\in(0,\\infty)$ im Punkt $z\\in\\Z^d$. Ziel ist es, ein Prinzip großer Abweichungen zu entwickeln, d.h. das Hauptaugenmerk liegt auf dem asymptotischen Verhalten der Wahrscheinlichkeit, dass die Selbstüberschneidungslokalzeiten von ihrem Erwartungswert in erheblichem Maße nach oben abweichen. Mit anderen Worten; es soll das asymptotische Verhalten von $$ \\log\\P(\\|\\ell_t\\|_p^p\\geq r^p_t) $$ genau bestimmt werden, wobei $r_t^p\\in(0,\\infty)$ schneller als der Erwartungswert $\\E[\\|\\ell_t\\|_p^p]$ gegen unendlich streben soll. Dieses Verhalten kann dabei durch $t$, $r_t$ und eine gewisse Variationsformel beschrieben werden. Es wird sich herausstellen, dass es zwei Fälle zu betrachten gilt, in denen sich das probabilistisch beste Verhalten stark unterscheidet; die genaue Position des Phasenübergangs hängt dabei von den Parametern $p$ und $d$ ab. Im Vorgriff auf die Resultate kann man festhalten, dass die nötigen Selbstüberschneidungen in kleinen Dimensionen (im sogenannten subkritischen Fall) über einen großen Bereich erfolgen, aufgrund dessen bei der mathematischen Modellierung eine Reskalierung erforderlich ist. In hohen Dimensionen (dem sogenannten superkritischen Fall) ist dies nicht nötig, da die erforderlichen Selbstüberschneidungen innerhalb eines begrenzten Intervalles erfolgen. Das Interesse an der Untersuchung entstand unter anderem aus der Verbindung zu Modellen der statistischen Mechanik (parabolisches Anderson Modell) und zur Variationsanalysis. In der Vergangenheit wurde eine Vielzahl an Methoden benutzt, um dieses Problem zu lösen. In der vorliegenden Dissertation soll die sogenannte Momentenmethode bestmöglich ausgereizt werden und es wird gezeigt, welche Ergebnisse damit möglich sind.

Vis didėjantis nusikaltimų skaičius veda prie vienos labai svarbios išvados – per menkai sprendžiamos socialinės problemos, o teisės aktų taikymas turi rimtų spragų. Visuomenės palaikymas labai svarbus asmenų socialinei adaptacijai gr��žus iš įkalinimo vietų, tai padeda buvusio nuteistojo jaustis pilnateisiais piliečiais, turinčiais eilę teisių kaip ir kiekvienas mūsų: teisę i darbą, mokslą, būstą, socialinę apsaugą. Vykdant socialinės paramos, pagalbos ir paslaugų teikimą vietiniame lygyje plačiau bendradarbiauti su nevyriausybinėmis organizacijomis, siekti socialinių paslaugų įvairovės, įtraukti į bendrą socialinės paramos sistemą rizikos grupės asmenims, sustiprinti jų ryšius visuomenėje. / The constantly growing number of crimes leads to one highly important conclusion: social problems are being poorly solved and the application of legal acts has serious gaps in our country. Public support is of high importance for the social adaptation of persons returning from prisons. This helps former convicts to feel as full citizens, having the same rights as other people: a right for work, education, lodgment and social security. It is necessary to cooperate with nongovernmental organizations closer, to strive for the diversity of social services, to include the persons belonging to the group disposed to risk into the general social support systems and to consolidate their links with society when implementing social support, assistance and service provision on local level.

This dissertation examines two main research questions: Why do firms deviate from their predicted level of toxic emissions, and how do these differences relate to financial performance? The objective is threefold: (1) to understand deviation in corporate environmental performance by looking at both industry and firm level variables, (2) to see how this deviation relates to both profitability and fluctuations in financial performance, and (3) to see if, and how, corporate environmental legitimacy affects the relationship between corporate environmental deviation and corporate financial performance. To achieve this objective the construct “corporate environmental performance deviation” is developed. It is defined as the extent to which a firm’s environmental performance deviates from its predicted performance, and is used to capture within-firm strategic choices in environmental management. Predicted environmental performance is calculated based on certain firm characteristics such as size and industry. Actual environmental performance is calculated using a weighted score of air emissions obtained from the Toxic Release Inventory (TRI) database. The difference between these two values represents a corporation’s environmental performance deviation. Corporate environmental performance deviation focuses on strategic choices related to environmental management, while recognizing that environmental management is the result of both institutional pressures and within-firm strategic decisions. Aligned with this focus, variables 2 related to this strategic choice are used to explain deviation in environmental management, including an environmental integration capability, firm strategy, and industry munificence and dynamism. Associated with the internal and external organizational analysis, institutional theory and the resource-based view (RBV) are used to explore the tension between deviation to increase competitiveness versus isomorphism to attain legitimacy. The sample is composed of 311 U.S. firms who have reported their toxic air releases to the TRI from 1998-2007. The sample is broken down into two subsets, those that exceed (positive deviation) or fail to meet (negative deviation) predicted environmental performance. Results of a longitudinal analysis show that positive environmental deviation is related to a greater capacity to strategically integrate environmental issues into a firm’s existing business approach, less munificence and dynamism in the task environment, and reduced financial fluctuations. Negative environmental deviation is decreased through a demonstrated capacity to strategically integrate environmental issues into a firm’s existing strategic approach, and related to greater munificence and dynamism in the task environment, reduced profitability and increased financial fluctuations. Lastly, although there are no significant main effects for corporate environmental legitimacy, the paradoxical combination of negative deviation and environmental legitimacy can reduce the severity of the negative financial results to negative deviation, both in terms of profitability and financial fluctuations.

This thesis investigates how to train the increasingly large cast of characters in modern commercial computer games. Modern computer games can contain hundreds or sometimes thousands of non-player characters that each should act coherently in complex dynamic worlds, and engage appropriately with other non-player characters and human players. Too often, it is obvious that computer controlled characters are brainless zombies portraying the same repetitive hand-coded behaviour. Commercial computer games would seem a natural domain for reinforcement learning and, as the trend for selling games based on better graphics is peaking with the saturation of game shelves with excellent graphics, it seems that better artificial intelligence is the next big thing. The main contribution of this thesis is a novel style of utility function, group utility functions, for reinforcement learning that could provide automated behaviour specification for large numbers of computer game characters. Group utility functions allow arbitrary functions of the characters’ performance to represent relationships between characters and groups of characters. These qualitative relationships are learned alongside the main quantitative goal of the characters. Group utility functions can be considered a multi-agent extension of the existing programming by reward method and, an extension of the team utility function to be more generic by replacing the sum function with potentially any other function. Hierarchical group utility functions, which are group utility functions arranged in a tree structure, allow character group relationships to be learned. For illustration, the empirical work shown uses the negative standard deviation function to create balanced (or equal performance) behaviours. This balanced behaviour can be learned between characters, groups and also, between groups and single characters. Empirical experiments show that a balancing group utility function can be used to engender an equal performance between characters, groups, and groups and single characters. It is shown that it is possible to trade some amount of quantitatively measured performance for some qualitative behaviour using group utility functions. Further experiments show how the results degrade as expected when the number of characters and groups is increased. Further experimentation shows that using function approximation to approximate the learners’ value functions is one possible way to overcome the issues of scale. All the experiments are undertaken in a commercially available computer game engine. In summary, this thesis contributes a novel type of utility function potentially suitable for training many computer game characters and, empirical work on reinforcement learning used in a modern computer game engine.

När en ny produkt lanseras på marknaden uppkommer det med stor sannolikhet brister som intetidigare upptäckts. Syftet med denna studie är att skapa en arbetsmodell för hur medelstora företagska hantera kvalitetsavvikelser som uppkommer inom produktion. För att skapa en realistisk modellför hur dessa kvalitetsavvikelser skall hanteras, baseras detta examensarbete på en fallstudie utfördpå ett tillverkande företag vid namn AQ ParkoPrint AB som är beläget i Gävle. Följande forskningsfrågor har besvarats för att understödja det huvudsakliga syftet: • Vilka kvalitetsavvikelser kan uppkomma inom medelstora företag? • Hur kan dessa kvalitetsavvikelser hanteras samt om möjligt reduceras? Metoderna som använts i studien är intervjuer och observationer. Fem intervjuer och tvåobservationer utfördes på AQ ParkoPrint AB. Den arbetsmodell som skapats kallas förfemstegsmodellen och innebär att processen för att hantera kvalitetsavvikelser delats upp i fem steg:dokumentera, utred, illustrera, forma åtgärdsförslag och genomför. Femstegsmodellen är utformadför medelstora tillverkande företag och inriktad på att skapa ett arbetssätt för systematiskt arbetemed hantering samt om möjligt reducering av kvalitetsbrister. Dessa steg är flexibla på det viset attvarje företag utformar egna åtgärdshanteringar och använder femstegsmodellen som grundstruktur idet fullständiga kvalitetsarbetet.Med hjälp av den utvecklade femstegsmodellen kan medelstora företag hantera och reducera sinakvalitetsbrister. Arbetsmodellen i sig kräver inga kostsamma verktyg utan det som styr kostnadernaär omfattningen av det arbete som läggs på samtliga steg.I denna studie presenteras även frågeställningar som företag kan arbeta utifrån för att reducerakvalitetsbrister som kan härledas till underleverantörer. Fallstudien påvisar att de kvalitetsavvikelsersom uppkommer troligtvis kan härledas till underleverantörerna. / When a new product is launched, product defects in various processes and process steps willprobably appear that were not previously known or detected. The purpose of this study is to create aworking model for how middle-sized companies should manage to handle defects in quality thatwill appear in manufacturing productions. To be able to create a realistic model, this thesis will bebased on a case study of a manufacturing company by the name of AQ ParkoPrint AB, located in Gävle, Sweden. The following research questions are answered in order to support the main purpose: • What defects in quality occur most often in middle-sized manufacturing companies? • How can these defects in quality be managed, and if possible, be reduced? The methods used in this study are interviews and observations. Five interviews and twoobservations were performed on AQ ParkoPrint AB. The working model that was created wasnamed “Five-step-model” and involves the process of managing defects in quality divided into fivesteps: document, investigate, illustrate, propose action and implement. The “Five-step model” isdesigned for middle-sized manufacturing companies focused on creating a way to worksystematically with how to managing quality defects and how to reduce them if possible. The stepsin this model are flexible because every single company should be able to customize each step tosuit their company needs, and only use the “Five-step model” that is created in this study as thebasic structure of the quality work. The case study shows that the suppliers are the majorcontributors of causing quality defects. Middle-sized companies can use the developed five stepmodel to handle and reduce their quality defects. The “Five-step model” itself requires no expensivetools. It is the extent of the work in every specific step that determines the costs.This study also presents some question formulations that companies can use to reduce their qualitydefects that can be traced back to the subcontractors.

I denna studie är syftet att undersöka och analysera personer som haft ett alkohol- och/eller narkotikaberoende, när det gäller deras uppfattning och upplevelse av omgivningens syn på dem och deras dåvarande missbruk. Vidare vill vi belysa och analysera om och i så fall på vilket sätt denna syn påverkat dem, i förhållande till deras missbruk. För att besvara syftet har följande frågeställningar använts: Hur upplever människor som har haft ett alkohol- och/eller narkotikaberoende att deras omgivning såg på dem och deras dåvarande missbruk? Vad var det avgörande för att dessa människor tog sig ur sitt beroende? Hur kan detta förstås ur ett stigmatiseringsperspektiv? Idag hävdar många att problemet med alkohol- och narkotikamissbruk inte är själva drogerna i sig utan att alltfler människor lever under ett negativt samhälleligt tryck. Via stämpling skapar detta tryck personliga besvär såsom ångest, självdestruktivitet och dålig självbild. Detta leder till ett vidare normstridigt beteende, som i sin tur leder till ytterligare stämplingsreaktioner. Dessa reaktioner förstärker problemen och resulterar i att människor försätts i en negativ och avvikande spiral. Rapporten vilar på en kvalitativ ansats med fenomenologi som vetenskapsteoretisk utgångspunkt. Vi söker alltså inte att förklara det sociala fenomen som undersöks utan snarare förstå det. Studiens datainsamlingsmetod var en kombination av halvstrukturerade och narrativa intervjuer. För att analysera vår insamlade data använde vi oss av meningsanalys som lägger fokus vid meningen samt betydelsen av våra egna antaganden och tolkningar. Vi har därmed blivit medskapare av intervjuernas och texternas mening. Resultatet av studien visar att människor med alkohol- och narkotikamissbruk kan uppleva att omgivningens syn på dem har betydelse för deras avvikande beteende.  Det visar också att dessa människor har blivit stämplade av andra och att de därför har en dålig självbild. Vidare visar resultatet att gemenskap är en förutsättning för att människor med alkohol- och narkotikamissbruk ska kunna förbättra sin självbild, ta sig ur sitt missbruk och fortsätta leva ett drogfritt liv / In this study, the purpose is to examine and analyze the experiences of people who have had an alcohol and/or drug addiction. What are their perception and experience of the social environment’s views on them and their abuse? Further, it is to highlight and discuss if and how these attitudes affects them. To answer that, the following questions have been used: How do people who have had an alcohol and/or drug addiction experience the way people in their surrounding look at them and their abuse? What was crucial for these people when it comes to get out of their addiction? How can this be understood from a stigma perspective? Today, many claims that the problem of alcohol and drug abuse is not just the drugs, but that an increasing amount of people are living under negative social pressure. Via stigmatization this creates personal inconveniences such as anxiety, self-destructiveness and a poor selfimage. This leads to an even more developed conflict according to traditional norms and behaviors, which in turn leads to further stigma reactions. These reactions strengthens the problems and results in an even more negative and deviant spiral. The investigastion has a qualitative design based on an phenomenological approach. The method for data collection is a combination of semi-structured and narrative interviews. To analyze our empirical results, we use the method of purpose analysis which focuses on the meaning and importance of ur own assumptions and interpretations. We have thus become co-creators of the interviews and the meaning of the texts. The result of the study shows that people with alcohol and drug addiction may feel that negative views on them have an impact for their continued deviant behavior. It also shows that these people have been stigmatized by others and that they therefore have a poor selfimage. Furthermore, the result shows that fellowship is a prerequisite for the possibility of people with alcohol and drug addiction to improve their self-image, to be able to abandon their addiction and continue to live a life without drugs

This dissertation examines two main research questions: Why do firms deviate from their predicted level of toxic emissions, and how do these differences relate to financial performance? The objective is threefold: (1) to understand deviation in corporate environmental performance by looking at both industry and firm level variables, (2) to see how this deviation relates to both profitability and fluctuations in financial performance, and (3) to see if, and how, corporate environmental legitimacy affects the relationship between corporate environmental deviation and corporate financial performance. To achieve this objective the construct “corporate environmental performance deviation” is developed. It is defined as the extent to which a firm’s environmental performance deviates from its predicted performance, and is used to capture within-firm strategic choices in environmental management. Predicted environmental performance is calculated based on certain firm characteristics such as size and industry. Actual environmental performance is calculated using a weighted score of air emissions obtained from the Toxic Release Inventory (TRI) database. The difference between these two values represents a corporation’s environmental performance deviation. Corporate environmental performance deviation focuses on strategic choices related to environmental management, while recognizing that environmental management is the result of both institutional pressures and within-firm strategic decisions. Aligned with this focus, variables 2 related to this strategic choice are used to explain deviation in environmental management, including an environmental integration capability, firm strategy, and industry munificence and dynamism. Associated with the internal and external organizational analysis, institutional theory and the resource-based view (RBV) are used to explore the tension between deviation to increase competitiveness versus isomorphism to attain legitimacy. The sample is composed of 311 U.S. firms who have reported their toxic air releases to the TRI from 1998-2007. The sample is broken down into two subsets, those that exceed (positive deviation) or fail to meet (negative deviation) predicted environmental performance. Results of a longitudinal analysis show that positive environmental deviation is related to a greater capacity to strategically integrate environmental issues into a firm’s existing business approach, less munificence and dynamism in the task environment, and reduced financial fluctuations. Negative environmental deviation is decreased through a demonstrated capacity to strategically integrate environmental issues into a firm’s existing strategic approach, and related to greater munificence and dynamism in the task environment, reduced profitability and increased financial fluctuations. Lastly, although there are no significant main effects for corporate environmental legitimacy, the paradoxical combination of negative deviation and environmental legitimacy can reduce the severity of the negative financial results to negative deviation, both in terms of profitability and financial fluctuations.

A survey of the literature of grinding and surface texture shows the influence of dressing and wear on surfaces involved in the process and the advantages of stylus profilometry for data collection from both grinding wheels and ground surfaces. Statistical analysis is favoured for surface profile characterization and, of the various parameters used, power spectral density alone offers some prospect of effective comparison between these surfaces. Work on grinding with single crystals of natural corundum was eventually discontinued in favour of experiments with conventional bonded grinding wheels subjected to a dressing operation and some wear in grinding steel surfaces. Statistical parameters representing the surfaces are computed using data obtained from profilograms. Results in terms of power spectral density are presented showing progressive improvement following upon developments in apparatus and methods which facilitated the use of larger surface profile samples. Transfer functions are used to relate power spectra representing corresponding pairs of surfaces. The significance of power spectral density applied to surface profile characterization is discussed and, in this context, it is suggested that these should be described as variance spectra. Attention is drawn to certain disadvantages of variance spectra applied to grinding wheel and ground surface profiles. Methods designed to improve presentation of variance spectra lead to development of a proposed new and more suitable spectrum in which density of standard deviation of surface profile ordinates with respect to frequency is plotted against frequency. Transfer functions calculated from related pairs of these standard deviation spectra show a strong linear correlation with frequency and offer prospects of convenient comparison between the profiles of the various surfaces involved in grinding.