Current Issue : October - December Volume : 2019 Issue Number : 4 Articles : 5 Articles
In this paper, we develop and parallelize a CFD solver that supports overlapped meshes on multiple MIC architectures by using\nmultithreaded technique. We optimize the solver through several considerations including vectorization, memory arrangement,\nand an asynchronous strategy for data exchange on multiple devices. Comparisons of different vectorization strategies are made,\nand the performances of core functions of the solver are reported. Experiments show that about 3.16x speedup can be achieved for\nthe six core functions on a single Intel Xeon Phi 5110P MIC card, and 5.9x speedup can be achieved using two cards compared to\nan Intel E5-2680 processor for two ONERA M6 wings case....
Just-in-time software defect prediction (JIT-SDP) is an active topic in software defect prediction, which aims to identify defect inducing\nchanges. Recently, some studies have found that the variability of defect data sets can affect the performance of defect\npredictors. By using local models, it can help improve the performance of prediction models. However, previous studies have\nfocused on module-level defect prediction. Whether local models are still valid in the context of JIT-SDP is an important issue. To\nthis end, we compare the performance of local and global models through a large-scale empirical study based on six open-source\nprojects with 227417 changes.The experiment considers three evaluation scenarios of cross-validation, cross-project-validation,\nand time wise-cross-validation. To build local models, the experiment uses the k-medoids to divide the training set into several\nhomogeneous regions. In addition, logistic regression and effort-aware linear regression (EALR) are used to build classification\nmodels and effort-aware prediction models, respectively. The empirical results show that local models perform worse than global\nmodels in the classification performance. However, local models have significantly better effort-aware prediction performance\nthan global models in the cross-validation and cross-project-validation scenarios. Particularly, when the number of clusters k is set\nto 2, local models can obtain optimal effort-aware prediction performance. Therefore, local models are promising for effort aware\nJIT-SDP....
The increasement of software complexity directly results in the augment of software fault and costs a lot in the process of software\ndevelopment and maintenance. The complex network model is used to study the accumulation and accumulation of faults in\ncomplex software as a whole.Then key nodes with high fault probability and powerful fault propagation capability can be found,\nand the faults can be discovered as soon as possible and the severity of the damage to the system can be reduced effectively. In\nthis paper, the algorithm MFS AN (mining fault severity of all nodes) is proposed to mine the key nodes fromsoftware network. A\nweighted software networkmodel is built by using functions as nodes, call relationships as edges, and call times asweight. Exploiting\nrecursive method, a fault probability metric FP of a function, is defined according to the fault accumulation characteristic, and a\nfault propagation capability metric FPC of a function is proposed according to the fault propagation characteristic. Based on the\nFP and FPC, the fault severity metric FS is put forward to obtain the function nodes with larger fault severity in software network.\nExperimental results on two real software networks show that the algorithm MFS _AN can discover the key function nodes correctly\nand effectively....
An essential objective of software development is to locate and fix defects\nahead of schedule that could be expected under diverse circumstances. Many\nsoftware development activities are performed by individuals, which may lead\nto different software bugs over the development to occur, causing disappointments\nin the not-so-distant future. Thus, the prediction of software defects\nin the first stages has become a primary interest in the field of software\nengineering. Various software defect prediction (SDP) approaches that rely\non software metrics have been proposed in the last two decades. Bagging,\nsupport vector machines (SVM), decision tree (DS), and random forest (RF)\nclassifiers are known to perform well to predict defects. This paper studies\nand compares these supervised machine learning and ensemble classifiers on\n10 NASA datasets. The experimental results showed that, in the majority of\ncases, RF was the best performing classifier compared to the others....
In order to improve software reliability, software defect prediction is applied to the process of software maintenance to identify\npotential bugs. Traditional methods of software defect prediction mainly focus on designing static code metrics, which are input\ninto machine learning classifiers to predict defect probabilities of the code. However, the characteristics of these artificial metrics\ndo not contain the syntactic structures and semantic information of programs. Such information is more significant than manual\nmetrics and can provide a more accurate predictive model. In this paper, we propose a framework called defect prediction via\nattention-based recurrent neural network (DP-ARNN). More specifically, DP-ARNN first parses abstract syntax trees (ASTs) of\nprograms and extracts them as vectors. Then it encodes vectors which are used as inputs of DP-ARNN by dictionary mapping and\nword embedding. After that, it can automatically learn syntactic and semantic features. Furthermore, it employs the attention\nmechanism to further generate significant features for accurate defect prediction. To validate our method, we choose seven opensource\nJava projects in Apache, using F1-measure and area under the curve (AUC) as evaluation criteria. The experimental results\nshow that, in average, DP-ARNN improves the F1-measure by 14% and AUC by 7% compared with the state-of-the-art\nmethods, respectively....
Loading....