Software defect prediction is one of the most popular research topics in software engineering. It aims to predict defect-prone modules before defects are discovered, therefore it can be used better prioritise quality assurance effort. In recent years, especially for 3 many new studies have been proposed. The goal this study comprehensively review, analyse and discuss state-of-the-art prediction. authors survey almost 70 representative papers years (January 2014–April 2017), which published...

Person re-identification has been widely studied due to its importance in surveillance and forensics applications. In practice, gallery images are high-resolution (HR) while probe usually low-resolution (LR) the identification scenarios with large variation of illumination, weather or quality cameras. this kind scenarios, which we call super-resolution (SR) person re-identification, not well studied. paper, propose a semi-coupled low-rank discriminant dictionary learning (SLD <sup...

Learning an expressive representation from multi-view data is a key step in various real-world applications. In this paper, we propose semi-supervised deep discriminant learning (SMDDRL) approach. Unlike existing joint or alignment methods that cannot simultaneously utilize the consensus and complementary properties of to learn inter-view shared intra-view specific representations, SMDDRL comprehensively exploits as well learns both representations by employing network. ignore redundancy...

With the expansion of data, increasing imbalanced data has emerged. When imbalance ratio (IR) is high, most existing learning methods decline seriously in classification performance. In this paper, we systematically investigate highly problem, and propose an uncorrelated cost-sensitive multiset (UCML) approach for it. Specifically, UCML first constructs multiple balanced subsets through random partition, then employs feature (MFL) to learn discriminant features from constructed multiset. To...

Person re-identification has been widely studied due to its importance in surveillance and forensics applications. In practice, gallery images are high resolution (HR), while probe usually low (LR) the identification scenarios with large variation of illumination, weather, or quality cameras. this kind scenarios, which we call super-resolution (SR) person re-identification, not well studied. paper, propose a semi-coupled low-rank discriminant dictionary learning (SLD2L) approach for SR task....

Heterogeneous defect prediction (HDP) refers to predicting defect-proneness of software modules in a target project using heterogeneous metric data from other projects. Existing HDP methods mainly focus on instances with single source. In practice, there exist plenty external Multiple sources can generally provide more information than project. Therefore, it is meaningful investigate whether the performance be improved by employing multiple sources. However, precondition conducting that are...

Video-based person re-identification (re-id) is an important application in practice. Since large variations exist between different pedestrian videos, as well within each video, it's challenging to conduct videos. In this paper, we propose a simultaneous intra-video and inter-video distance learning (SI2DL) approach for video-based re-id. Specifically, SI2DL simultaneously learns intravideo metric from the training The used make video more compact, one ensure that truly matching videos...

Web service recommendation plays an important role in building service-oriented systems. QoS-based has recently gained much attention for providing a promising way to help users find high-quality services. To accurately predict the QoS values of candidate services, systems usually need collect historical data from users, which will potentially pose threat user's privacy. However, how simultaneously protect privacy and make accurate prediction not been well studied. By taking these two...

Person re-identification plays an important role in video surveillance and forensics applications. In many cases, person needs to be conducted between image clip, e.g., re-identifying a suspect from large quantities of pedestrian videos given single the suspect. We call this scenario as reidentification (IVPR). practice, are usually represented with different features, there exist variations frames within each video. These factors make matching become very challenging task. paper, we propose...

Heterogeneous defect prediction (HDP) aims to predict defect-prone software modules in one project using heterogeneous data collected from other projects. Recently, several HDP methods have been proposed. However, these do not sufficiently incorporate the two characteristics of data: (1) could be linearly inseparable, and (2) highly imbalanced. These make it challenging build an effective model. In this paper, we propose a novel Ensemble Multiple Kernel Correlation Alignment (EMKCA) based...

Video-based person re-identification (re-id) has attracted a lot of research interest. When facing dramatic growth in new pedestrian videos, existing video-based re-id methods usually need large quantities labeled videos to train discriminative model. In practice, labeling is costly and time-consuming task, which will limit the application these real environment. Therefore, it valuable necessary investigate how learn model by using limited training videos. this paper, we propose...

Language bias stands as a noteworthy concern in visual question answering (VQA), wherein models tend to rely on spurious correlations between questions and answers for prediction. This prevents the from effectively generalizing, leading decrease performance. In order address this bias, we propose novel modality fusion collaborative de-biasing algorithm (CoD). our approach, is considered model’s neglect of information particular during We employ training approach facilitate mutual modeling...

Summary In the early phases of software testing, projects may have only limited historical defect data. Learning prediction model with such insufficient training data will limit efficacy learned predictor. practice, there are usually many publicly available fault datasets. Recently, heterogeneous (HFP) has been proposed. However, existing HFP models do not investigate how to use mixed project predict target. Furthermore, often imbalanced. The imbalanced distribution source leads serious...