No description available.

No description available.

No description available.

No description available.

No description available.

Deep learning models heavily rely on an abundance of data, and their performance is directly affected by data availability. In mobility pattern modeling, problems, such as next location prediction or flow prediction, are commonly solved using deep learning approaches. Despite advances in modeling techniques, complications arise when acquiring mobility data is limited by geographic factors and data protection laws. Generating highquality synthetic data is one of the solutions to get around at times when information is scarce. Trajectory generation is concerned with generating trajectories that can reproduce the spatial and temporal characteristics of the underlying original mobility patterns. The task of this project was to evaluate Generative Adversarial Network (GAN) capabilities to generate synthetic vehicle trajectory data. We extend the methodology of previous research on trajectory generation by introducing conditional trajectory duration labels and a model pretraining mechanism. The evaluation of generated trajectories consisted of a two-fold analysis. We perform qualitative analysis by visually inspecting generated trajectories and quantitative analysis by calculating the statistical distance between synthetic and original data distributions. The results indicate that extending the previous GAN methodology allows the novel model to generate trajectories statistically closer to the original data distribution. Nevertheless, a statistical base model has the best generative performance and is the only model to generate visually plausible results. We accredit the superior performance of the statistical base model to the highly predictive nature of vehicle trajectories, which must follow the road network and have the tendency to follow minimum distance routes. This research considered only one type of GAN-based model, and further research should explore other architecture alternatives to understand the potential of GAN-based models fully / Modeller för djupinlärning är starkt beroende av ett överflöd av data, och derasprestanda påverkas direkt av datatillgänglighet. I mobilitetsmönstermodellering löses problem, såsom nästa platsförutsägelse eller flödesprediktion,vanligtvis med hjälp av djupinlärningsmetoder. Trots framsteg inommodelleringsteknik uppstår komplikationer när inhämtning av mobilitetsdatabegränsas av geografiska faktorer och dataskyddslagar. Att generera syntetiskdata av hög kvalitet är en av lösningarna för att ta sig runt i tider dåinformationen är knapp. Bangenerering handlar om att generera banorsom kan reproducera de rumsliga och tidsmässiga egenskaperna hos deunderliggande ursprungliga rörlighetsmönstren. Uppgiften för detta projektvar att utvärdera GAN-kapaciteten för att generera syntetiska fordonsbanor. Viutökar metodiken för tidigare forskning om banagenerering genom att introducera villkorliga etiketter för banalängd och en modellförträningsmekanism.Utvärderingen av genererade banor bestod av en tvåfaldig analys. Viutför kvalitativ analys genom att visuellt inspektera genererade banor ochkvantitativ analys genom att beräkna det statistiska avståndet mellan syntetiskaoch ursprungliga datafördelningar. Resultaten indikerar att en utvidgningav den tidigare GAN-metoden tillåter den nya modellen att generera banorstatistiskt närmare den ursprungliga datadistributionen. Ändå har en statistiskbasmodell den bästa generativa prestandan och är den enda modellen somgenererar visuellt rimliga resultat. Vi ackrediterar den statistiska basmodellensöverlägsna prestanda till den mycket prediktiva karaktären hos fordonsbanor,som måste följa vägnätet och ha en tendens att följa minimiavståndsrutter.Denna forskning övervägde endast en typ av GAN-baserad modell, ochytterligare forskning bör utforska andra arkitekturalternativ för att förståpotentialen hos GAN-baserade modeller fullt ut

Persistent conduct problems, which include aggressive, disruptive, and antisocial behaviors, are associated with a range of negative correlates and sequelae, including school failure and dropout, depression, and substance use. Problematic peer relationships, such as peer rejection and deviant peer affiliation (DPA), have been implicated in both the development and persistence of conduct problems. In particular, the impact of peer rejection has been highlighted in childhood, whereas the effects of DPA have received more attention during adolescence. Conduct problems, peer rejection, and DPA are dynamic processes that change across development, and understanding the relations among these processes is particularly important among contextually disadvantaged youth who are at elevated risk for aggression. Nevertheless, limited research has considered both peer rejection and DPA as they relate to conduct problems in a single study and even less research has looked at potential moderators of the relations among these processes. To address these gaps, I empirically evaluated a developmental model of conduct problems and peer processes, focusing on individual differences across childhood and adolescence. The goal of the present study was to examine continuity and discontinuity of trajectories of conduct problems and problematic peer relations from childhood through adolescence and to identify contextual factors that impact continuity/discontinuity of risk. Participants were 678 urban children recruited in fall of 1st grade for the second generation Prevention Intervention Research Center's Periodic Follow-up of Two Preventative Interventions. Aggression and peer rejection in childhood were modeled separately first, followed by modeling the joint trajectories of aggression and rejection. Four distinct joint trajectories of aggression-rejection in grades 1 to 3 were identified: high decreasing aggression-high stable rejection, low increasing aggression-moderate increasing rejection, moderate decreasing aggression-moderate decreasing rejection, and low stable aggression-low stable rejection. As with the childhood analyses, aggression and DPA in adolescence were modeled separately first, followed by modeling the joint trajectories of aggression and DPA. Three distinct joint trajectories of aggression-deviant peer affiliation in grades 6 to 12 were identified: high decreasing aggression-low stable DPA, moderate stable aggression-high stable DPA, and low stable aggression-low stable DPA. Next, latent transition patterns from high- and low-risk childhood aggression-rejection trajectories to adolescent aggression-DPA trajectories pointed to evidence for both continuity and discontinuity of risk along aggression-DPA trajectories. Introducing contextual variables (i.e., perceived neighborhood adversity, community violence exposure, and poor parental monitoring) into the transition model suggested that higher levels of contextual risk were associated with elevated risk for transition to higher-risk adolescent aggression-DPA classes (i.e., high decreasing aggression-low stable DPA, moderate stable aggression-high stable DPA). / Psychology

This research effort aims to develop a series of full-sized humanoid robots, and to research a simple but reliable bipedal walking method. Since the debut of Wabot from Waseda University in 1973, several full-sized humanoid robots have been developed around the world that can walk, and run. Although various humanoid robots have successfully demonstrated their capabilities, bipedal walking methods are still one of the main technical challenges that robotics researchers are attempting to solve. It is still challenging because most bipedal walking methods, including ZMP (Zero Moment Point) require not only fast sensor feedback, but also fast and precise control of actuators. For this reason, only a small number of research groups have the ability to create full-sized humanoid robots that can walk and run. However, if we consider this problem from a different standpoint, the development of a full-sized humanoid robot can be simplified as long as the bipedal walking method is easily formulated. Therefore, this research focuses on developing a simple but reliable bipedal walking method. It then presents the designs of two versions of a new class of super lightweight (less than 13 kg), full-sized (taller than 1.4 m) humanoid robots called CHARLI-L (Cognitive Humanoid Autonomous Robot with Learning Intelligence – Lightweight) and CHARLI-2. These robots have unique designs compared to other full- sized humanoid robots. CHARLI-L utilizes spring assisted parallel four-bar linkages with synchronized actuation to achieve the goals of lightweight and low cost. Based on the experience and lesions learned from CHARLI-L, CHARLI-2 uses gear train reduction mechanisms, instead of parallel four-bar linkages, to increase actuation torque at the joints while further reducing weight. Both robots successfully demonstrated untethered bipedal locomotion using an intuitive walking method with sinusoidal foot movement. This walking method is based on the ZMP method. Motion capture tests using six high speed infrared cameras validate the proposed bipedal walking method. Additionally, the total power and energy consumptions during walking are calculated from measured actuator currents. / Ph. D.

Self-thinning quantifies the reduction in tree numbers due to density-dependent mortality. Maximum size-density relationships (MSDRs) are a component of self-thinning that describe the maximum tree density per unit area obtainable for a given average tree size, often quadratic mean diameter (D). An MSDR species boundary line has been defined as a static upper limit of maximum tree density -- D relationships that applies to all stands of a certain species within a particular geographical area. MSDR dynamic thinning lines have been defined as the maximum tree density obtainable within an individual stand for a particular D which have been shown to vary relative to planting density. Results from this study show that differences in boundary levels of individual stands cause the MSDR species boundary line slope estimate to be sensitive to the range of planting densities within the model fitting dataset. Thus, a second MSDR species boundary line was defined whose slope is the average slope of all MSDR dynamic thinning lines. Mixed-models are presented as a statistical method to obtain an estimate of the population average MSDR dynamic thinning line slope. A common problem when modeling self-thinning is to determine what observations are within generally accepted stages of stand development. Segmented regression is presented as a statistical and less subjective method to determine what observations are within various stages of stand development. Estimates of D and trees per acre (N) where MSDR dynamic thinning lines begin and end on the logarithmic scale were used as response variables and predicted as a function of planting density. Predictions of MSDR dynamic thinning line beginning and ending D and N are used in an alternative MSDR dynamic thinning line slope estimation method. These models show that the maximum value of Reineke's Stand Density Index (SDI) varies relative to planting density. By relating planting density specific Zone of Imminent Competition Mortality boundaries to a MSDR species boundary line, self-thinning was found not to begin at a constant relative SDI. Thus, planting density specific Density Management Diagrams (DMD) showed that self-thinning began at 40 to 72% for planting densities of 605 and 2722 seedlings per acre, respectively. / Ph. D.

The problem of developing a generalized impulse as a function of a set of parameters is investigated. The proposed generalized impulse alters an existing orbit by producing, over some period of time, a change in velocity, ΔV, as well as a change in position, Δr. The generalized impulse is described by parameters associated with an instantaneous change in velocity as well as parameters associated with an atmospheric skip trajectory. Closed form solutions are obtained through several changes of independent variable, the use of modified Chapman variables and the consequent analytical integration of the uncoupled equations. The closed form solutions contain between two and six parameters depending on the complexity of the desired skip trajectory. Fuel optimal transfer orbits are obtained using the generalized impulse along with Keplerian arcs and instantaneous changes in velocity. Families of coplanar and noncoplanar transfers for circular orbit to circular orbit are numerically generated. The generated transfer trajectories involve the rendezvous of two vehicles. The orbits are not globally optimal but rather optimal for the specified number and type of velocity impulses specified. The optimal solution to the nonlinear problem is determined via sequential quadratic programming which satisfies the Kuhn-Tucker optimality conditions for constrained minimization. It is found that for transfer between coplanar and noncoplanar orbits, solutions using the generalized impulse compare favorably with solutions obtained by optimal control theory. Numerical solution to complex problems involving transfer from general orbit to general orbit were not obtained. / Master of Science