Abstract. Robust Point Matching (RPM) is a common image registration algo-rithm, yet its large computational complexity prohibits registering large point sets in a timely manner. With recent advances in General Purpose Graphical Process-ing Units (GPGPUs), commodity hardware is capable of greatly reducing the execution time of RPM when non-rigidly aligning thousands of data points. In this paper, we identify areas where parallelism can be exploited in the RPM algo-rithm, and investigate a GPU-based approach to accelerate the implementation. Other common RPM implementations are compared with our solution. Experi-ments on synthetic and real data sets show that our approach achieves close to linear speed-up with respect to total computational power over the widely used Matlab implementation. Our tests indicate that utilizing our implementation on current state of the art GPU technology would enable the use of vastly greater point set sizes. 1

Abstract: Using data from a survey of 160 urban borrowing groups of the Microfund for Women in Jordan, we investigate the effect of screening, peer monitoring, group pressure, and social ties on borrowing groups ’ repayment behavior as an indirect test of different theoretical models. The dependent variable used captures the intensity of default measured by the total number of days of late repayment after each due date, allowing us to use count data models with cluster standard errors. As theory predicts, our empirical analysis suggests that peer monitoring, group pressure, and social ties reduce delinquency. The paper uncovers interesting evidence about the role of social ties and religion. Most notably, in an area where religion contributes to attitudes and beliefs of individuals, we find that religiosity improves repayment performance.

In the medical imaging field, we need fast deformable registration methods especially in intra-operative settings characterized by their time-critical applications. Image registration studies which are based on Graphics Processing Units (GPUs) provide fast implementations. However, only a small number of these GPU-based studies concentrate on deformable registration. We implemented Demons, a widely used deformable image registration algorithm, on NVIDIA’s Quadro FX 5600 GPU with the Compute Unified Device Architecture (CUDA) programming environment. Using our code, we registered 3D CT lung images of patients. Our results show that we achieved the fastest runtime among the available GPU-based Demons implementations. Additionally, regardless of the given dataset

Does the group leader matter? The impact of monitoring activities and social ties of group leaders on the repayment performance of group-based lending in Eritrea

In this paper, we build on the important study of Guiso, Sapienza and Zingales (2004) on the role of social capital in promoting financial access, and make several extensions to the literature. First, ours is the first paper to consider the impact of social capital on access to credit in a developing country setting, where the problems of financial access are clearly more pronounced than in developed or middle-income countries. Second, we use both individual and aggregate-level measures of social capital, whereas the previous literature has used only aggregate measures. Third, we extend the literature of financial access to consider issues of organizational choice. We test a variety of hypotheses by using a nationally representative, individual-level survey of financial access. Our results indicate that individual social capital is positively and significantly associated with access to credit from financial institutions, whereas we do not find a significant relationship between general interpersonal trust and access to institutional loans. Our results for organizational choice indicate that the importance of individual social capital as a borrower screening device increases when the formality of the financial institution decreases.

The last fifty years have seen the large-scale implementation of financial programs that specifically target the poor. Many such early programs, often implemented by governments, were not able to satisfactorily deal with information uncertainties, and hence resulted in weak incentive structures, heavily bureaucratic and politicized approaches and inevitably low repayment performance. More recently (even though the concept is more than a century old), a new type of approach to lending to the poor—known as microfinance—has designed specific methods to deal with information uncertainties, resulting in impressive program performance. Almost concurrently, social scientists and development practitioners have identified and assembled growing anecdotal evidence to suggest that the forms of capital traditionally used in growth theory (natural, physical and human) are missing an important element. This concept, generically known as “social capital”, includes the various networks of relationships among economic actors, and the values and attitudes associated with them. A large part of the success of microfinance programs resides in their ability to surmount the significant information problems inherent in dealing with poor customers with no banking experience and unknown creditworthiness. Hence this literature review examines how social capital can help reduce the cost of imperfect information in small financial transactions, and thereby improve the performance of credit delivery programs in