Driving as a human: a track learning based adaptable architecture for a car racing controller

We present the evolution and current state of the Mr. Racer car racing controller that excelled at the corresponding TORCS competitions of the last years. Although several heuristics and black-box optimization methods are employed, the basic idea of the controller architecture has been to take over many of the mechanisms human racing drivers apply. They learn the track geometry, plan ahead, and wherever necessary, adapt their plans to the current circumstances quickly. Mr. Racer consists of several modules that have partly been adapted and optimized separately, where the final tuning is usually done with respect to a certain racing track during the warmup phase of the TORCS competitions. We also undertake an experimental evaluation that investigates how the controller profits from adding some of the modules to a basic configuration and which modules are most important for reaching the best possible performance.     

Mesh-free canonical tensor products for six-dimensional density matrix: computation of kinetic energy

The computation of a six-dimensional density matrix is the crucial step for the evaluation of kinetic energy in electronic structure calculations. For molecules with heavy nuclei, one has to consider a very refined mesh in order to deal with the nuclear cusps. This leads to high computational time and needs huge memory for the computation of the density matrix. To reduce the computational complexity and avoid discretization errors in the approximation, we use mesh-free canonical tensor products in electronic structure calculations. In this paper, we approximate the six-dimensional density matrix in an efficient way and then compute the kinetic energy. Accuracy is examined by comparing our computed kinetic energy with the exact computation of the kinetic energy.     

The Sadeian interface: computers and catharsis

This paper explores the limitations of contemporary interface design and offers the potential of more profound forms of interaction by drawing on the rich and much older heritage of interactive art. Whilst HCI design is preoccupied with making the computer more simple to use, installation work, kinetic sculpture, and interactive multimedia art forms have generally been more concerned with the predicament of human/technological negotiation, whilst remaining a salient form of human communication. HCI activity sets out to make the complex systems of computing easy to understand and use, whilst interactive art often uses simple technology to make complex, inspiring and esoteric statements and experiences. In many ways the more simple and 'low resolution' the technology the more immersive, acute and intimate the experience. 'Low resolution' examples such as telephone-sex-lines are explored alongside more immersive systems, such as biofeedback interfaces, and other interactive experiments drawn from the 'technic' strand of art history.     

Drought Severity Assessment Based on Bivariate Probability Analysis

Conventionally drought severity is assessed based on drought indices. Recently the Reconnaissance Drought Index (RDI) was proposed to assess drought severity based on the precipitation to potential evapotranspiration ratio (P/PET). In this paper RDI is studied as a bivariate index under a set of assumptions and simplifications. The paper presents a simple computational procedure for estimating the P/PET ratio for selected reference periods varying from 3 to 12 months, for any return period of drought. Alternatively, based on this procedure, the severity of any drought episode is rationally assessed. A bivariate probability analysis is employed based on the assumption that P and PET values are normally distributed and often negatively correlated. Examples for the application of the proposed procedure are presented using data from several meteorological stations in Greece. It is shown that the assumption of normality of both P and PET holds for long periods at all examined stations.     

CPU demand for web serving: Measurement analysis and dynamic estimation

Managing the resources in a large Web serving system requires knowledge of the resource needs for service requests of various types. In order to investigate the properties of Web traffic and its demand, we collected measurements of throughput and CPU utilization and performed some data analyses. First, we present our findings in relation to the time-varying nature of the traffic, the skewness of traffic intensity among the various types of requests, the correlation among traffic streams, and other system-related phenomena. Then, given such nature of web traffic, we devise and implement an on-line method for the dynamic estimation of CPU demand.

Assessing resource needs is commonly performed using techniques such as off-line profiling, application instrumentation, and kernel-based instrumentation. Little attention has been given to the dynamic estimation of dynamic resource needs, relying only on external and high-level measurements such as overall resource utilization and request rates. We consider the problem of dynamically estimating dynamic CPU demands of multiple kinds of requests using CPU utilization and throughput measurements. We formulate the problem as a multivariate linear regression problem and obtain its basic solution. However, as our measurement data analysis indicates, one is faced with issues such as insignificant flows, collinear flows, space and temporal variations, and background noise. In order to deal with such issues, we present several mechanisms such as data aging, flow rejection, flow combining, noise reduction, and smoothing. We implemented these techniques in a Work Profiler component that we delivered as part of a broader system management product. We present experimental results from using this component in scenarios inspired by real-world usage of that product.     

Coupling agent-based models of subsistence farming with individual-based forest models and dynamic models of water distribution

Subsistence farming communities are dependent on the landscape to provide the resource base upon which their societies can be built. A key component of this is the role of climate and the feedback between rainfall, crop growth, land clearance and their coupling to the hydrological cycle. Temporal fluctuations in rainfall alter the spatial distribution of water availability, which in turn is mediated by soil-type, slope and landcover. This pattern ultimately determines the locations within the landscape that can support agriculture and controls sustainability of farming practices. The representation of such a system requires us to couple together the dynamics of human and ecological systems and landscape change, each of which constitutes a significant modelling challenge on its own. Here we present a proto-type coupled modelling system to simulate land-use change by bringing together three simple process models: (a) an agent-based model of subsistence farming; (b) an individual-based model of forest dynamics; and (c) a spatially explicit hydrological model which predicts distributed soil moisture and basin scale water fluxes. Using this modelling system we investigate how demographic changes influence deforestation and assess its impact on forest ecology, stream hydrology and changes in water availability.     

iPhone改变PC世界?

下一代台式PC和笔记本电脑也许会跟iPhone特别相像。一年前,我写过一篇专栏文章＂为什么iPhone会改变PC世界＂,在那篇文章中我描绘了未来PC操作系统的用户界面,将出现类似于iPhone中的多重触控、智能识别、3D和微小图标等元素。在微软的Surface Computing中这些元素已经成为现实。     

Optimization problems in contracted tensor networks

We discuss the calculus of variations in tensor representations with a special focus on tensor networks and apply it to functionals of practical interest. The survey provides all necessary ingredients for applying minimization methods in a general setting. The important cases of target functionals which are linear and quadratic with respect to the tensor product are discussed, and combinations of these functionals are presented in detail. As an example, we consider the representation rank compression in tensor networks. For the numerical treatment, we use the nonlinear block Gauss?CSeidel method. We demonstrate the rate of convergence in numerical tests.     

Automatically performing weak mutation with the aid of symbolic execution,concolic testing and search-based testing

Automating software testing activities can increase the quality and drastically decrease the cost of software development. Toward this direction, various automated test data generation tools have been developed. The majority of existing tools aim at structural testing, while a quite limited number aim at a higher level of testing thoroughness such as mutation. In this paper, an attempt toward automating the generation of mutation-based test cases by utilizing existing automated tools is proposed. This is achieved by reducing the killing mutants’ problem into a covering branches one. To this extent, this paper is motivated by the use of state of the art techniques and tools suitable for covering program branches when performing mutation. Tools and techniques such as symbolic execution, concolic execution, and evolutionary testing can be easily adopted toward automating the test input generation activity for the weak mutation testing criterion by simply utilizing a special form of the mutant schemata technique. The propositions made in this paper integrate three automated tools in order to illustrate and examine the method’s feasibility and effectiveness. The obtained results, based on a set of Java program units, indicate the applicability and effectiveness of the suggested technique. The results advocate that the proposed approach is able to guide existing automating tools in producing test cases according to the weak mutation testing criterion. Additionally, experimental results with the proposed mutation testing regime show that weak mutation is able to speedup the mutant execution time by at least 4.79 times when compared with strong mutation.