Earticle

Predicting crime types by using classification algorithms can help to find factors affecting crimes and prevent crimes. Due to various reasons in the process of data collection, there are often a large number of missing values in actual criminal dataset, which seriously affects the classification accuracy. Therefore, based on mutual KNNI (K nearest neighbor imputation) algorithm and combined with GRA (Grey Relational Analysis) theory, a novel data filling algorithm called GMKNN is proposed in order to improve the classification accuracy. The algorithm replaces the Euclidean distance formula used in mutual KNNI algorithm with the Grey relational grade formula to eliminate the effect of noise from the nearest neighbors and effectively deal with the discrete attributes. By comparing with several popular data filling algorithms based on a real criminal dataset with lots of missing values, higher classification accuracy can be obtained by using GMKNN algorithm, which is up to 77.837%.

In this paper, we present a robust and fast tracking method based on joint collaborative representation. Traditional sparse coding based tracking methods code the candidates as a sparse linear combination of a series of object and trivial templates and perform time consuming L1 regularizations. In contrast to these methods, this paper adopts the L2-regularized least square models to reduce the computational complexity. The tracked object can be represented by the linear combination of a series of object templates, and also can be represented by candidate samples in the current frame. We propose a joint objective function to handle the tracking process. In addition, we introduce an effective update scheme to deal with the change of target appearance over time. Experiments on several challenging image sequences show that our proposed tracking method is robust and efficient.

Coal structure not only is a significant physical parameter of the coal reservoirs, but also has an effect on the coalbed methane (CBM) productivity during coal mining. In this paper, the clustering method is used to evaluate the coal structures quantitatively. For better application on the identification process of the coal structures, the clustering method is modified twice by ascertaining initial clustering centers and calibrating Euclidean distances, respectively. The quantitative method was applied to evaluate coal structures in 16 CBM wells in the Zhengzhuang field, Southern Qinshui Basin, North China. Results show that coal structures by the twice modified clustering method has good accordance with that by the coring samples. Therefore, coal structure identification using clustering method with the aid of geophysical logging data is feasible in the study area. The thickness proportions of coal structures have positive/ negative relationships with the corresponding average logging data. The multiple linear regression method was employed to predict the thickness proportions of coal structures. The predicted results show that prediction models of thickness proportions of coal structures are feasible to predict the distribution of coal structures fast in spite of the limited errors.

In the present work, Electrical discharge machining (EDM) was employed along with silicon abrasive powders mixed in dielectric fluid to machine AISI 52100 tool steel workpiece with copper electrode by adopting centered central composite (CCD) design of response surface methodology. To realize the process performance of material removal rate and surface roughness variables such as Pulse on time, duty cycle, peak current and the Silicon powder concentration with particle concentration and size of 2.5–2.8 g/L and 45–50μm, respectively, have been added into the kerosene dielectric liquid of a die-sinking electrical discharge machine. Analysis of variance (ANOVA) has been applied to study the influence of process parameters and their interactions. In addition a mathematical model has been formulated in order to estimate the process performance. Further, the parameters were optimized for maximizing MRR and minimizing SR using desirability function approach. The optimum settings of parameters are pulse on time 200μ𝑠, duty cycle 0.81, peak current 11A, concentration 4 g/l, for maximizing MRR and minimizing SR.The recommended optimal process conditions have been verified by conducting confirmation experiments. The obtained optimal settings are experimentally verified showing +5.2% and -4.65% as the relative errors for MRR and SR respectively.

Software reliability growth models (SRGMs) investigating software debugging process have recently been developed by some researchers to estimate software reliability measures. However, these SRGMs either did not involve resource expenditures spent on fault remove process or assumed that the rate of resources consumed is constant. In practice, software fault intensity may be not continuous for many conditions. Thus, in order to address the problem, this paper develops models incorporating the resource expenditures and change-point, which spent on software debugging process. A real software failure project is demonstrated the effectiveness of proposed models, and numerical results prove new models can provide better fit and prediction.

Spectral clustering is a method of subspace clustering which is suitable for the data of any shape and converges to global optimal solution. By combining concepts of shared nearest neighbors and geodesic distance with spectral clustering, a self-adaptive spectral clustering based on geodesic distance and shared nearest neighbors was proposed. Experiments show that the improved spectral clustering algorithm can fully take into account the information of neighbors, but also measure the exact distance and better process the geodetic data.

During any system's development life cycle, software testing is considered an important and key phase. Even so, the testing phase is commonly the first phase to be removed when time and resources constraints are faced due to its high consumption of recourses. As a result, developers waste their time and effort on developing software that is either filled with bugs or mismatches user specifications. The excessive work and time spent to program and execute a test case takes approximately 70% of a project's resources. The goal of this work is to produce a practical implementation of the extreme programming as a software development methodology that focuses on generating test cases from the design information before initiating the coding phase. Furthermore, a competent testing tool is built to employ the available information in the related design diagrams, Class and State diagram, wherein it automatically generates transition sequences to test any software system. The proposed tool can reduce the time, cost, and human resources used to test or redesign the system when it mismatches user specifications and requirements.