Magíster en Tecnologías de la Información / Un problema recurrente en un sistema es la degradación del rendimiento a través del tiempo, las variaciones suelen ser en un principio casi imperceptibles hasta que llega un punto de inflexión donde el aumento de los tiempos de respuesta se hace cada vez mayor. En este punto la identificación y corrección de las desviaciones en el comportamiento esperado del sistema, se vuelve crítico ya que dado el comportamiento exponencial en el aumento de los tiempos de respuesta, nos pone contra el tiempo para evitar el colapso del sistema. Lo anterior pone en riesgo la continuidad operacional de la organización, la posible disrupción del servicio puede traer consecuencias económicas para la compañía y nuestros clientes, dada esta situación se hace crítico corregir las desviaciones de manera oportuna, rápida y a un bajo costo, sin embargo, no contamos con una forma de identificar que piezas de software son las responsables de los problemas de performance detectados. Se propone el concepto de Vertical Profiling, la cual es una técnica y metodología que fue definida en la Universidad de Lugano, la cual utilizaremos para generar una representación del comportamiento de una aplicación basado en un set de métricas, donde cada métrica se representa como una serie de tiempo, esto nos permite entender el comportamiento a través del tiempo en cada capa del sistema, independiente de su nivel de abstracción y/o tipo de componente. Esto implica 2 etapas, la primera que consiste en un proceso de recolección de datos a través de todas las capas que componen el software y la segunda etapa que consiste en poder hacer un análisis que considere la variación en el tiempo del comportamiento del sistema. Para poder realizar el análisis del comportamiento, se propone una visualización gráfica, multidimensional, de fácil lectura y oportuna, con el fin de poder identificar los cuellos de botella. Para este estudio en particular, se utilizará la orientación a aspectos para la captura de datos, y el análisis se realizará utilizando Moose, la cual es una plataforma de análisis de datos, en este caso lo emplearemos para el análisis de Software, y para la visualización utilizaremos Mondrian que nos permite crear el meta-modelo y así realizar un análisis visual del rendimiento del aplicativo. Con la implementación de nuestro Vertical Profiling, se redujeron considerable los tiempos de respuesta de nuestra aplicación de forma rápida, algo que con un profiler tradicional no fue posible.

<p>As the embedded devices has become more and more generalized and as their product cycles keeps shrinking the field has opened up for the Application Specific Instruction set Processor. A mix between the classic generalized microcontroller and the specialized ASIC the ASIP keeps a set of general processing instructions for executing embedded software but combines that with a set of heavily specialized instructions for speeding up the data intense application core algorithms. One important aspect of the ASIP design flow</p><p>research is cutting design time and cost. One way of that is automation of the instruction set design. In order to do so a process is needed where the algorithm to be ASIPed is analyzed and critical operations are found and exposed so that they can be implemented in special hardware. This process is called profiling. This thesis describes an implementation of a fine grained source code profiler for use in an ASIP design flow. The profiler software is based on a static-dynamic workflow where data is assembled from both static</p><p>analysis and dynamic execution of the program and then analyzed together in an specially made analysis software.</p>

Thesis (Ph.D.)--RAND Graduate School, 2007. / Includes bibliographical references.

Racial profiling, in the context of the current study, concerns the association of racial and/or ethnic status with criminality and manifests in the traffic stop. The body of knowledge now available on racial profiling has documented well the incidence of numerical disparity of traffic stops between racial groups, with motorists of color subject to intrusion by the state at greater rates than White motorists (Withrow 2005). Criminologists then turned to ‘perception’-based research to examine what makes an individual ‘perceive’ he has been racially profiled. I argue that the second wave of research is dominated by a narrow survey approach, concentrates on the microlevel police-citizen encounter, and lacks a theoretical grounding, particularly in race theory. The ‘perception’ orientation, I argue, discursively diminishes the experiences of communities of color in their experiences with the state. The current study re-examines the two main components of the ‘perception’ based research -- personal and vicarious experience with the police – to extend our understanding of the meanings behind personal and vicarious encounters with law enforcement. The current qualitative study, based on more than two dozen in-depth interviews, informs our understanding of racial profiling on a number of levels. Citizenship emerges as a dominant narrative from my respondents, thus extending the effects of the racialized traffic stop effects beyond the particularistic police-minority relationship and into larger legal and political realms not anticipated in the current literature. I find that the ‘shadow citizenship’ identity imposed by the state through racializing and criminalizing processes like racial profiling is regularly rejected by people of color through various forms of resistance to racial oppression. A third important finding concerns the complication of ‘vicarious experiences.’ My respondents indicate that they do not summarily adopt views about the police but contextualize their own experiences within understandings of collective memory. Finally, because I engage racial profiling through the theoretical perspectives of Collins, DuBois, Feagin, and Foucault, among others, and frame my overall research approach using critical race theory, the salience of race in racial profiling processes is undeniably evident, contrary to the racial vacuum dominating the current literature.

The aim of this thesis was to investigate different aspects of male breast cancer (MBC), and to compare these with findings in female breast cancer (FBC). In paper I, a population–based study was performed to investigate possible differences in treatment and outcome between MBC and FBC patients. MBC and FBC presented with a similar distribution of stage. Although no differences in primary treatment strategy were demonstrated, MBC patients had significantly poorer overall and relative survival, indicating a more aggressive disease. Paper II aimed to assess the value of clinicopathological factors and molecular subtypes in MBC. One hundred and ninety-seven MBC tumors were characterized using immunohistochemistry (IHC) and the findings were correlated to outcome. Lymph node positivity, larger tumor size and ER-negativity were independent risk factors for breast cancer death. Tumor grade, HER2, Ki 67 or IHC classification into molecular subtypes did not demonstrate any prognostic information. In paper III, the same patient material as in paper II was used for evaluation of proliferation markers. High levels of cyclin A and cyclin B expression and an elevated mitotic count were predictive of breast cancer death. Ki-67 was re-evaluated using different cut-offs, but no prognostic value could be demonstrated. Contrarily, overexpression of cyclin D1 was associated with a lower risk of breast cancer death. In papers IV-V, the molecular background of MBC tumors was investigated.  Global GEX analyses were performed and two novel subgroups of MBC tumors were identified; luminal M1 and luminal M2. When comparing the degree of similarity with the “intrinsic” subtypes in FBC tumors, more than half of the MBC tumors remained unclassified.  Comparative genomic hybridization was used to investigate DNA aberrations. Two MBC subgroups were identified, of which one did not resemble any of the female subgroups. In both studies on the molecular level, a majority of patients were classified into the subgroup with a more aggressive tumor behavior. In conclusion, MBC seems to be a unique tumor entity. The established molecular subtypes in FBC are not applicable in MBC. Other prognostic profiles, specific for MBC, need to be identified.

Rock grouting is performed to decrease the hydraulic conductivity around underground structures, such as tunnels and caverns. Cement grouts are often used and pumped into joint and fractures of the rock formation. Piston type pumps are mostly used for high pressure rock grouting. A pulsation effect is inevitable when using this type of pump due to the movement of the piston. The effect of this pulsation on rock grouting is yet to be known but believed to be benefi-cial for the penetration of the grout. Current flow meters used in the field are not accu-rate enough to determine the fluctuation of the flow rate when it is less than 1 l/min. In addition, currently available flow meters measure the average of the flow over a cer-tain period of time, hence the true fluctuation of the flow rate due to the pulsation of the piston remains unknown. In this paper, a new methodology, the so called ‘Ultrasound Velocity Profiling – Pressure Difference’ (UVP+PD) method has been introduced to show the pulsation effect when using a piston type pump. The feasibility of this method was successfully investigated for the direct in-line determination of the rheological properties of micro cement based grouts under field conditions (Rahman &amp; Håkansson, 2011). Subse-quently, it was also found that this method can be very efficient to measure the fluctu-ation of the flow rate for different types of pumps. From a grouting point of view the UVP+PD method can be used to synchronize the pressure and flow of a piston type pump by measuring the pulsation effect. Conse-quently it can be used as a tool for the efficiency and quality control of different types of pumps. / <p>QC 20121221</p>

Abstract INDIVIDUAL-TECHNOLOGY FIT: MATCHING INDIVIDUAL CHARACTERISTICS AND FEATURES OF BIOMETRIC INTERFACE TECHNOLOGIES WITH PERFORMANCE By ADRIANE B. RANDOLPH MAY 2007 Committee Chair: Dr. Melody Moore Jackson Major Department: Computer Information Systems The term biometric literally means “to measure the body”, and has recently been associated with physiological measures commonly used for personal verification and security applications. In this work, biometric describes physiological measures that may be used for non-muscularly controlled computer applications, such as brain-computer interfaces. Biometric interface technology is generally targeted for users with severe motor disabilities which may last long-term due to illness or injury or short-term due to temporary environmental conditions. Performance with a biometric interface can vary widely across users depending upon many factors ranging from health to experience. Unfortunately, there is no systematic method for pairing users with biometric interface technologies to achieve the best performance. The current methods to accommodate users through trial-and-error result in the loss of valuable time and resources as users sometimes have diminishing abilities or suffer from terminal illnesses. This dissertation presents a framework and methodology that links user characteristics and features of biometric interface technologies with performance, thus expediting the technology-fit process. The contributions include an outline of the underlying components of capturing and representing individual user characteristics and the impact on the performance of basic interaction tasks using a methodology called biometric user profiling. In addition, this work describes a methodology for objectively measuring an individual’s ability to control a specific biometric interface technology such as one based on measures of galvanic skin response or neural activity. Finally, this work incorporates these concepts into a new individual-technology fit framework for biometric interface technologies stemming from literature on task-technology fit. Key words: user profiles, biometric user profiling, biometric interfaces, fit, individual-technology fit, galvanic skin response, functional near-infrared, brain-computer interface

Offender profiling aims at generating information of an offender based upon the crime committed and the goal is to be an adjunct in the crime invesigation process. The aim of the study was to examine crime scene characteristics and corroborating evidence and their associations with offenders' interrogation behavior. Results from 207 analyzed police files indicated that reactive (impulsive) offenders confessed more often when there was technical evidence against them compared to instrumental (calculating) offenders. Reactive offenders also had more injuries. Injured reactive offenders were more inclined to deny, whereas injured instrumental offenders claimed amnesia more often. When comparing the confessions injured reactive offenders were more inclined to confess than injured instrumental offenders. There was also a strong tendency implying more confessions with witness reports. The findings point towards the practical usefulness of information of crime scene characteristics and corroborating evidence in the interrogation phase by better tailoring the interviewing strategy.

This thesis presents a protocol for intervehicle communication for use in Vehicular Ad Hoc Networks (VANET). VANET is a natural extension of mobile ad hoc networks (MANET) in which the restrictions related to power and mobility are relaxed. The routing protocols used for MANETs are generally dependent on the state of the network. With changes in the network topology, routing messages are generated so that the states of the routers in the network are updated. In the case of VANETs, in which the level of node mobility is high, message-routing overhead has serious implications for the scalability and throughput of the routing protocol. This thesis introduces criteria that are recommended for use when protocols are designed for VANET applications and presents the Selected Node Geographic Forwarding (SNGF) protocol. The SNGF protocol implements controlled flooding in an efficient manner in order to reduce unnecessary communication overhead. The protocol has a destination discovery mechanism that allows it to initiate correspondence between nodes without reliance on static location services. The protocol avoids formation of clusters by using the concept of selective forwarding, thus providing the advantages of cluster based approaches without actually forming one itself. It effectively deals with blind flooding by introducing a comprehensive retransmission time delay in the nodes. This retransmission delay favors the nodes in the direction of the destination and prevents other nodes from retransmitting the same message. The SNGF protocol does not use routing tables, which require frequent updates in mobile networks, instead it relies on directing the messages to geographic locations which are forwarded by any available intermediary nodes. The protocol also provides techniques for handling network fragmentation which can be a frequent problem in vehicular networks. It is capable of delayed message transmission and multiple route discovery in the case of the non-availability of the shortest path to the destination. To evaluate the performance of the SNGF protocol, an extensive study of mobile networks was conducted using the NS2 simulator. The simulation results demonstrate the reachability of the protocol, its scalability advantages and its total independence from location services. The SNGF protocol allows each participating node to operate independently of other nodes in the network. Nodes in the network are able to communicate with other nodes without ever becoming dependent on intermediary nodes. This feature opens new possibility for individual node based application development in ad hoc networks. The traffic profiling is described as it would be observed by an independent node participating in VANET using the SNGF protocol. The node communicates with other nodes and collects relevant data through the discourse capability of SNGF. The data collected by the node is viewed as a snapshot in time of the traffic conditions down the road based upon which future traffic condition is predicted. Traffic profiling is investigated for different levels of VANET deployment. The simulation results show that the proposed method of traffic profiling in a VANET environment using the SNGF protocol is viable for even lower levels of deployment.

Racial profiling, in the context of the current study, concerns the association of racial and/or ethnic status with criminality and manifests in the traffic stop. The body of knowledge now available on racial profiling has documented well the incidence of numerical disparity of traffic stops between racial groups, with motorists of color subject to intrusion by the state at greater rates than White motorists (Withrow 2005). Criminologists then turned to ‘perception’-based research to examine what makes an individual ‘perceive’ he has been racially profiled. I argue that the second wave of research is dominated by a narrow survey approach, concentrates on the microlevel police-citizen encounter, and lacks a theoretical grounding, particularly in race theory. The ‘perception’ orientation, I argue, discursively diminishes the experiences of communities of color in their experiences with the state. The current study re-examines the two main components of the ‘perception’ based research -- personal and vicarious experience with the police – to extend our understanding of the meanings behind personal and vicarious encounters with law enforcement. The current qualitative study, based on more than two dozen in-depth interviews, informs our understanding of racial profiling on a number of levels. Citizenship emerges as a dominant narrative from my respondents, thus extending the effects of the racialized traffic stop effects beyond the particularistic police-minority relationship and into larger legal and political realms not anticipated in the current literature. I find that the ‘shadow citizenship’ identity imposed by the state through racializing and criminalizing processes like racial profiling is regularly rejected by people of color through various forms of resistance to racial oppression. A third important finding concerns the complication of ‘vicarious experiences.’ My respondents indicate that they do not summarily adopt views about the police but contextualize their own experiences within understandings of collective memory. Finally, because I engage racial profiling through the theoretical perspectives of Collins, DuBois, Feagin, and Foucault, among others, and frame my overall research approach using critical race theory, the salience of race in racial profiling processes is undeniably evident, contrary to the racial vacuum dominating the current literature.