Cancer's heterogeneity necessitates precise subtype identification for effective diagnosis and treatment, which can be achieved by integrating multi-omics data to reveal distinct molecular characteristics and enable personalized therapies. Recently, significant efforts have been made through contrastive clustering methods to efficiently identify cancer subtypes. However, existing approaches remain limited in effectively capturing inter- and intra-view relationships in multi-omics data. Additionally, most cancer subtyping methods often rely on random sampling to construct negative pairs, which may inadvertently engender false negatives. To overcome these challenges, we propose a novel end-to-end self-supervised learning model named Decoupled Contrastive Multi-view Clustering with adaptive false negative elimination (DCMC). Specifically, DCMC adopts a multi-view clustering architecture that facilitates intra- and inter-view contrastive learning across distinct embedding spaces, allowing view-specific information to be preserved while maintaining cross-view consistency. We further introduce an adaptive false negative elimination framework to progressively screen potential false negatives. Finally, pseudo-label rectification is applied to enhance the quality of the learned representations and further refine the clustering process. DCMC is evaluated on 10 commonly used cancer datasets against 19 state-of-the-art methods, with experimental results validating its superior performance. In the Liver Hepatocellular Carcinoma case study, differential expression analysis is performed to identify potential biomarkers, while the cancer subtypes identified by DCMC are validated for their responses to specific therapeutic drugs. The datasets and source code for DCMC are available online at https://github.com/LinMengX/DCMC.

The miniaturization of multi-rotor aerial vehicles (MAVs) poses challenges to autonomous navigation. Cameras are preferred due to the limited payload of MAVs. In this paper, we present an optical flow (OF)-based approach for obstacle avoidance in indoor environments with a payload constrained MAV. Our approach mainly utilizes Time-to-Contact (TTC) and balance strategy (BS), two basic principles demonstrated effectively in many different obstacle avoidance systems. Free spaces through which a MAV can fly are estimated by TTC. BS is used to guide a MAV flying in a collision free path. Our approach is implemented in the framework of Robot Operating System (ROS). The simulation demonstrates the effectiveness of the approach.

Optical flow plays an important role in vision-based navigation. A lot of optical flow based methods have been developed for autonomous robots. In this paper, we evaluate obstacle avoidance methods based on optical flow in synthetic and real-world scenes. Specifically, the balance strategy is chosen and five representative optical flow algorithms are used. The new metrics for obstacle avoidance performance are introduced. Our experiments demonstrate the effectiveness of the balance strategy for obstacle avoidance. Furthermore, some factors contributing to obstacle avoidance ability are studied.

Dynamic topic models (DTM) are of great use toanalyze the evolution of unobserved topics of a text collectionover time. Recent years have witnessed the explosive growth ofstreaming text data emerging from online media, which createsan unprecedented need for DTMs for timely event analysis. While there have been some matrix factorization methods inthe literature for dynamic topic modeling, further study is stillin great need to model emerging, evolving and fading topicsin a more natural and effective way. In light of this, we firstpropose a matrix factorization model called SONMFSR (SoftOrthogonal NMF with Sparse Representation), which makes fulluse of soft orthogonal and sparsity constraints for static topicmodeling. Furthermore, by introducing the constraints of emerging, evolving and fading topics to SONMFSR, we easily obtain a novel DTM called SONMFSRd for dynamic event analysis. Extensive experiments on two public corpora demonstrate the superiority of SONMFSRd to some state-of-the-art DTMs in both topic detection and tracking. In particular, SONMFSRd shows great potential in real-world applications, where popular topics in Two Sessions 2015 are captured and traced dynamically for possible insights.

This paper presents a 3D frontier-based exploration tool named 3D-FBET. Our tool runs onboard the MAV equipped with a 3D sensor. The 3D map of the environment explored is constructed incrementally from two consecutive point clouds obtained. Considering the computation and memory limitations of MAVs, the OctoMap is utilized to represent 3D models. A novel approach is designed to extract the 3D frontiers from the OctoMap. Different from existing extraction method, only state-changed space in the 3D map is processed in each iteration. We implement our approach on top of the well-known robot operating system (ROS) and demonstrate the effectiveness of our tool in real scenarios.

Testing of image processing applications is a challenging job especially, when evaluating the correctness of output image. Generally, output images are evaluated manually by visual inspection carried out by an expert tester, which is the main hindrance in automation of testing process. Recently, statistical and metamorphic testing approaches are presented to automate output evaluation of image processing applications. The statistical method is dependent on availability of statistical distribution of output images, whereas metamorphic testing require more research efforts to make it widely used in practice. Metamorphic testing is a well-known technique to alleviate the test oracle problem and eliminates the required manual efforts by using relations of input and output images. Follow-up test cases are generated based on these relations and their expected output is evaluated. This paper addresses test oracle problem for image processing applications and demonstrates how properties of implementation under test can be adopted as metamorphic relations. We have studied general and specific metamorphic relations of morphological image operations such as dilation and erosion. Selection of metamorphic relations and their effectiveness by mutation analysis is demonstrated. The results show that metamorphic testing is useful for evaluation of output images in the absence of a perfect test oracle.

This paper presents a novel technique to generate test images using segmental symbolic evaluation for testing of image processing applications. Images are multidimensional and diverse in nature, which leads to different challenges for the testing process. A technique is required to generate test images capable of finding program paths derived by image pixels. The proposed technique is based on symbolic execution which is extensively used for test data generation in recent years. In image processing applications, pixel operations such as averaging, convolution etc. are applied on a segment of input image pixels called window for a single iteration and repeated for the entire image. Our key idea is to imitate operations on pixel window using symbolic values rather than concrete ones to generate path constraints in the program under test. The path constraints generated for different paths are solved for concrete values using our simple SAT solver and the solutions are capable to guide program execution to the specific paths. The solutions of path constraints are used to generate synthetic test images for each identified path and the paths constraints which are not solvable for concrete pixel values are reported as infeasible paths. We have developed a tool IMSUITthat takes an image processing function as input and executes the program symbolically for the given pixels window to generate test images. Effectiveness of IMSUIT is tested on different modules of an optical character recognition system and the result shows that it can successfully create test images for each path of the program under test and capable of identifying infeasible paths.

Software debugging starts with bug reports. Test engineers confirm bugs and determine the corresponding developers to fix them. However, the analysis of bug reports is time-consuming and manual inspection is difficult and tedious. To improve the efficiency of the whole process, we propose a bug processing framework that integrates bug report analysis and fault localization. An instance of the framework is implemented for regression faults. Preliminary results on a large open source application demonstrate both efficiency and effectiveness.

Script-based automated regression testing is widely used in industry. In this work, we focus on failed tests in a real regression test project. The causes of 197 failed tests produced in automated testing are examined and categorized based on an analysis procedure presented. The result shows that incorrect scripts, oracle mismatches, test tool bugs and misconfigurations involved in testing contribute most to failures instead of product bugs. Detecting and fixing the false positives are laborious and time-consuming. Our empirical study demonstrates that the benefits of test automation are obliterated to some extent in practical settings. To improve the effectiveness of automated regression testing further, some issues should receive more attention.

Regression faults are inevitably introduced in software development. Identifying and fixing regression faults can be tedious and time-consuming. The goal of my doctoral research is to provide an automated practical technique to effectively and efficiently locating failure-inducing changes. In this work, the research problem and related work is discussed first. Then, our approach, research questions, completed work and future work is presented. Finally, the expected contributions of my thesis are listed.

During software evolution, new released versions still contain many bugs. One common scenario is that end users encounter regression faults and submit them to bug tracking systems. Different from in-house regression testing, typically only one test input is available, which passes the old version and fails the modified new version. To address the issue, delta debugging has been proposed for failure-inducing changes identification between two versions. Despite promising results, there are two practical factors that thwart the application of delta debugging: a large number of tests and misleading false positives. In this work, we present a combination of coverage analysis and delta debugging that automatically isolates failure-inducing changes. Evaluations on twelve real regression faults in GNU software demonstrate both the speed gain and effectiveness improvements. Moreover, a case study on libPNG and TCPflow indicates that our technique is comparable to peer techniques in debugging regressions faults.

Open source projects often maintain open bug repositories during development and maintenance, and the reporters often point out straightly or implicitly the reasons why bugs occur when they submit them. The comments about a bug are very valuable for developers to locate and fix the bug. Meanwhile, it is very common in large software for programmers to override or overload some methods according to the same logic. If one method causes a bug, it is obvious that other overridden or overloaded methods maybe cause related or similar bugs. In this paper, we propose and implement a tool Rebug-Detector, which detects related bugs using bug information and code features. Firstly, it extracts bug features from bug information in bug repositories; secondly, it locates bug methods from source code, and then extracts code features of bug methods; thirdly, it calculates similarities between each overridden or overloaded method and bug methods; lastly, it determines which method maybe causes potential related or similar bugs. We evaluate Rebug-Detector on an open source project: Apache Lucene-Java. Our tool totally detects 61 related bugs, including 21 real bugs and 10 suspected bugs, and it costs us about 15.5 minutes. The results show that bug features and code features extracted by our tool are useful to find real bugs in existing projects.

Recent studies have demonstrated that language-based static analysis is capable of finding hundreds of bugs in complex real systems. Such static analysis allows users to specify properties in a specification language on demand. Paths in control flow graphs are explored exhaustively against user-defined properties. To avoid the potential path explosion problem, many techniques have been used in practice such as summaries. In this paper, we investigate how to simplify programs under check utilizing user-specified properties. From our observations, most functions under check are irrelevant to given properties. Checking those functions is time consuming. A program sifting approach is proposed to select functions related to properties. Sifters are derived from user-specified properties automatically. Functions matched or affected by sifters are safely preserved while the others are safely removed. We implemented a tool SIFT and carried out some experiments. Results show that SIFT is capable of simplifying program under check remarkably with small cost. In our experiments, 85% functions in files are sifted out while 89% analysis time is saved on average.