No description available.

No description available.

No description available.

No description available.

Data that is represented with high dimensionality presents a computational complexity challenge for many existing algorithms. Limiting dimensionality by discarding attributes is sometimes a poor solution to this problem because significant high-level concepts may be encoded in the data across many or all of the attributes. Non-linear dimensionality reduction (NLDR) techniques have been successful with many problems at minimizing dimensionality while preserving intrinsic high-level concepts that are encoded with varying combinations of attributes. Unfortunately, many challenges remain with existing NLDR techniques, including excessive computational requirements, an inability to benefit from prior knowledge, and an inability to handle certain difficult conditions that occur in data with many real-world problems. Further, certain practical factors have limited advancement in NLDR, such as a lack of clarity regarding suitable applications for NLDR, and a general inavailability of efficient implementations of complex algorithms. This dissertation presents a collection of papers that advance the state of NLDR in each of these areas. Contributions of this dissertation include: • An NLDR algorithm, called Manifold Sculpting, that optimizes its solution using graduated optimization. This approach enables it to obtain better results than methods that only optimize an approximate problem. Additionally, Manifold Sculpting can benefit from prior knowledge about the problem. • An intelligent neighbor-finding technique called SAFFRON that improves the breadth of problems that existing NLDR techniques can handle. • A neighborhood refinement technique called CycleCut that further increases the robustness of existing NLDR techniques, and that can work in conjunction with SAFFRON to solve difficult problems. • Demonstrations of specific applications for NLDR techniques, including the estimation of state within dynamical systems, training of recurrent neural networks, and imputing missing values in data. • An open source toolkit containing each of the techniques described in this dissertation, as well as several existing NLDR algorithms, and other useful machine learning methods.

No description available.

Sambandet mellan den fysiska miljön och brottslighet har sedan länge varit en relevant fråga inom den kriminologiska diskursen. Den föreliggande studien ämnar vidare undersöka huruvida fysisk oordning i urbana bostadsområden kan studeras genom Google Street View, ett webbaserat instrument för virtuella observationer. Syftet med studien är att undersöka om virtuellt observerad och självrapporterad uppfattad grad av fysisk oordning i bostadsområdet mäter samma fenomen, men även om virtuellt observerad fysisk oordning kan förklara skillnader i självrapporterad utsatthet för egendomsbrott. Genom att utföra virtuella observationer av fysisk oordning med hjälp av Google Street View i tjugo centralt belägna bostadsområden i Malmö visar resultaten att observerad och självrapporterad grad av fysisk oordning är starkt korrelerade och förefaller mäta samma fenomen. Resultaten visar även att observerad nivå av fysisk oordning genom Google Street View till viss del kan förklara variansen av utsatthet för egendomsbrott mellan bostadsområden. Avslutningsvis framhålls i studien att virtuella observationer genom Google Street View är ett lovande samt potentiellt kostnadseffektivt tillvägagångssätt för att undersöka graden av fysisk oordning i urbana bostadsområden. Användandet av Google Street View kantas dock av flera begränsningar som både framhålls och diskuteras grundligt i denna studie. / The correlation of physical environment and crime has been an ever relevant topic in the criminological discourse. This study attempts to unravel whether physical disorder in inner-city urban neighborhoods may be studied through Google Street View as a virtual observational tool. The aims of the study is to examine whether virtually observed and self-reported perceived level of neighborhood disorder measure the same phenomenon, and whether virtually observed physical disorder may explain variations of self-reported victimization of property crimes. By conducting virtual observations of physical disorder in twenty inner-city neighborhoods of Malmö through Google Street View, the results of the study propose that virtually observed and self-reported perceived level of disorder is strongly correlated and thus seems to measure the same phenomenon to a great extent. The results of the study also imply that observed physical disorder through Google Street View also accounts for neighborhood differences in victimization of property crimes. The study concludes that virtual observation through Google Street View is a promising and potentially cost-effective alternative approach when auditing neighborhood physical disorder. The methodology does however suffer by limitations which is highlighted and thoroughly discussed.

No description available.

The operational performance of an underground mine depends critically on how the production is scheduled. Increasingly advanced methods are used to create optimized long-term plans, and simultaneously the actual excavation is getting more and more automated. Therefore, the mapping of long-term goals into tasks by manual short-term scheduling is becoming a limiting segment in the optimization chain. In this thesis we study automating the short-term mine scheduling process, and thus contribute to an important missing piece in the pursuit of autonomous mining. First, we clarify the fleet scheduling problem and the surrounding context. Based on this knowledge, we propose a flow shop that models the mine scheduling problem. A flow shop is a general abstract process formulation that captures the key properties of a scheduling problem without going into specific details. We argue that several popular mining methods can be modeled as a rich variant of a k-stage hybrid flow shop, where the flow shop includes a mix of interruptible and uninterruptible tasks, after-lags, machine unavailabilities, and sharing of machines between stages. Then, we propose a Constraint Programming approach to schedule the underground production fleet. We formalize the problem and present a model that can be used to solve it. The model is implemented and evaluated on instances representative of medium-sized underground mines. After that, we introduce travel times of the mobile machines to the scheduling problem. This acknowledges that underground road networks can span several hundreds of kilometers. With this addition, the initially proposed Constraint Programming model struggles with scaling to larger instances. Therefore, we introduce a second model. The second model does not solve the interruptible scheduling problem directly; instead, it solves a related uninterruptible problem and transforms the solution back to the original time domain. This model is significantly faster, and can solve instances representative of large-sized mines even when including travel times. Lastly, we focus on finding high-quality schedules by introducing Large Neighborhood Search. To do this, we present a domain-specific neighborhood definition based on relaxing variables corresponding to certain work areas. Variants of this neighborhood are evaluated in Large Neighborhood Search and compared to using only restarts. All methods and models in this thesis are evaluated on instances generated from an operational underground mine. / Underjordsgruvans operativa prestanda är till stor del beroende av schemaläggningen av de mobila maskinerna. Allt mer avancerade metoder används för att skapa optimerade långtidsplaner samtidigt som produktionsaktiviteterna blir allt mer automatiserade. Att överföra långtidsmål till aktiviteter genom manuell schemaläggning blir därför ett begränsande segment i optimeringskedjan. I denna avhandling studerar vi automatisering av schemaläggning för underjordsgruvor och bidrar således med en viktig komponent i utvecklandet av autonom gruvdrift. Vi börjar med att klargöra schemaläggningsproblemet och dess omgivande kontext. Baserat på detta klargörande föreslår vi en abstraktion där problemet kan ses som en flow shop. En flow shop är en processmodell som fångar de viktigaste delarna av ett schemaläggningsproblem utan att hänsyn tas till allt för många detaljer. Vi visar att flera populära gruvbrytningsmetoder kan modelleras som en utökad variant av en k-stage hybrid flow shop. Denna utökade flow shop innehåller en mix av avbrytbara och icke avbrytbara aktiviteter, eftergångstid, indisponibla maskiner samt gemensamma maskinpooler för vissa steg. Sedan föreslår vi ett koncept för att lösa schemaläggningsproblemet med hjälp av villkorsprogrammering. Vi formaliserar problemet och presenterar en modell som kan användas för att lösa det. Modellen implementeras och utvärderas på probleminstanser representativa för mellanstora underjordsgruvor. Efter det introducerar vi restider för de mobila maskinerna i schemaläggningsproblemet. Detta grundar sig i att vägnätet i underjordsgruvor kan sträcka sig upp till flera hundra kilometer. Med det tillägget får den initiala villkorsprogrammeringsmodellen svårt att lösa större instanser. För att möta det problemet så introducerar vi en ny modell. Den nya modellen löser inte det avbrytbara problemet direkt utan börjar med att lösa ett relaterat, icke avbrytbart, problem för att sedan transformera lösningen tillbaka till den ursprungliga tidsdomänen. Denna modell är betydligt snabbare och kan lösa probleminstanser representativa för stora underjordsgruvor även när restider inkluderas. Avslutningsvis fokuserar vi på att hitta scheman av hög kvalitet genom att optimera med Large Neighborhood Search. För att åstadkomma detta presenterar vi ett domänspecifikt grannskap baserat på att relaxera variabler som rör aktiviteter inom vissa produktionsområden. Flera varianter av detta grannskap utvärderas och jämförs med att enbart använda omstarter. Alla metoder och modeller i den här avhandlingen är utvärderade på genererade instanser från en operativ underjordsgruva. / <p>QC 20181026</p>

Social disorganization theory proposes that neighborhood characteristics, such as residential instability, racial and ethnic heterogeneity, concentrated disadvantage, and immigrant concentration contribute to an increase in crime rates. Informal social controls act as a mediator between these neighborhood characteristics and crime and delinquency. Informal social controls are regulated by members of a community and in a disorganized community these controls are not present, therefore, crime and delinquency flourish (Sampson, 2012). Researchers have focused on these measures of social disorganization and the ability to explain a variety of crimes, specifically public crimes. Recently, researchers have focused their attention to characteristics of socially disorganized areas and the ability to predict private crimes, such as domestic violence. This study contributes to the research on social disorganization theory and domestic violence by examining domestic offenses at three different units of analysis: Chicago census tracts, Chicago neighborhoods, and Illinois counties. Demographic variables from the 2005-2009 American Community Survey were utilized to measure social disorganization within Chicago census tracts, Chicago neighborhoods, and Illinois counties. Data on domestic offenses in Chicago were from the City of Chicago Data Portal and data on domestic offenses in Illinois counties were retrieved from the Illinois Criminal Justice Information Authority (ICJIA). This study incorporated geographic information systems (GIS) mapping to examine the relationships between locations of domestic offenses and the measures of social disorganization in each unit of analysis. Results of this study indicate that different measures of social disorganization are significantly associated with domestic offenses in each unit of analysis.