Search

Deep neural networks are vulnerable to adversarial examples, even in the black-box setting where the attacker is only accessible to the model output. Recent studies have devised effective black-box attacks with high query efficiency. However, such performance is often accompanied by compromises in attack imperceptibility, hindering the practical use of these approaches. In this article, we propose to restrict the perturbations to a small salient region to generate adversarial examples that can hardly be perceived. This approach is readily compatible with many existing black-box attacks and can significantly improve their imperceptibility with little degradation in attack success rates. Furthermore, we propose the Saliency Attack, a new black-box attack aiming to refine the perturbations in the salient region to achieve even better imperceptibility. Extensive experiments show that compared to the state-of-the-art black-box attacks, our approach achieves much better imperceptibility scores, including most apparent distortion (MAD), L 0 and L 2 distances, and also obtains significantly better true success rate and effective query number judged by a human-like threshold on MAD. Importantly, the perturbations generated by our approach are interpretable to some extent. Finally, it is also demonstrated to be robust to different detection-based defenses.

Since automatic algorithm configuration methods have been very effective, recently there is increasing research interest in utilizing them for automatic solver construction, resulting in several notable approaches. For these approaches, a basic assumption is that the given training set could sufficiently represent the target use cases such that the constructed solvers can generalize well. However, such an assumption does not always hold in practice since in some cases, we might only have scarce and biased training data. This article studies effective construction approaches for the parallel algorithm portfolios that are less affected in these cases. Unlike previous approaches, the proposed approach simultaneously considers instance generation and portfolio construction in an adversarial process, in which the aim of the former is to generate instances that are challenging for the current portfolio, while the aim of the latter is to find a new component solver for the portfolio to better solve the newly generated instances. Applied to two widely studied problem domains, that is, the Boolean satisfiability problems (SAT) and the traveling salesman problems (TSPs), the proposed approach identified parallel portfolios with much better generalization than the ones generated by the existing approaches when the training data were scarce and biased. Moreover, it was further demonstrated that the generated portfolios could even rival the state-of-the-art manually designed parallel solvers.

The Dial-a-ride problem (DARP) is a challenging combinatorial optimization problem with lots of applications in the real world. It has so many applications in the smart city like bus intelligent scheduling and taxi intelligent scheduling. However, most of these works rarely address how to handle the infeasible initial solution. The infeasible solution is an unstable factor in the search framework which gives a negative influence on the quality of the final local optimal solution. In this paper, we present a heuristic repair method (HRM) for DARP. Based on a greedy strategy and proper evaluation function, the HRM will reduce the infeasibility of the most infeasible route by deleting several waypoints in the route and reinsert them into one feasible route. The result of experiments shows that HRM can raise about 50% of the fixing ability of a local search operator.

Social recommendations utilize social relations to enhance the representation learning for recommendations. Most social recommendation models unify user representations for the user-item interactions (collaborative domain) and social relations (social domain). However, such an approach may fail to model the users heterogeneous behavior patterns in two domains, impairing the expressiveness of user representations. In this work, to address such limitation, we propose a novel Disentangled contrastive learning framework for social Recommendations DcRec. More specifically, we propose to learn disentangled users representations from the item and social domains. Moreover, disentangled contrastive learning is designed to perform knowledge transfer between disentangled users representations for social recommendations. Comprehensive experiments on various real-world datasets demonstrate the superiority of our proposed model., Comment: CIKM2022

Cytochrome P450 (CYP), a multi-gene superfamily, is involved in a broad range of physiological processes, including hormone responses and secondary metabolism throughout the plant life cycle. Longan (Dimocarpus longan), a subtropical and tropical evergreen fruit tree, its embryonic development is closely related to the yield and quality of fruits. And a large number of secondary metabolites, such as flavonoids and carotenoids, are also produced during the longan somatic embryogenesis (SE). It is important, therefore, to study potential functions of CYPs in longan. However, the knowledge of longan CYPs is still very limited. Here, a total of 327 DlCYPs were identified using the genome-search method, which could be classified into nine clans. The expansion of the DlCYP family was mainly caused by tandem duplication (TD) events. Promoter cis-acting elements analysis elucidated that DlCYPs played important roles in hormonal responses. A total of 246 DlCYPs exhibited six different expression patterns during the early SE based on longan transcriptomic data. Eight DlCYPs underwent alternative splicing (AS) events, and they might produce one to six isoforms. And the AS transcript of DlCYP97C1 might act as an alternative to the full-length transcript in ICpEC and GE stages. Finally, protein-protein interaction (PPI) networks and miRNA target prediction elucidated that DlCYPs might be involved in the phenylpropanoid metabolic pathway and primarily regulated and targeted by miR413. In summary, our results provided valuable inventory for understanding the classification and biological functions of DlCYPs and provided insight into further functional verification of DlCYPs during the longan early SE.

The Travelling Salesman Problem (TSP) is a classical NP-hard problem and has broad applications in many disciplines and industries. In a large scale location-based services system, users issue TSP queries concurrently, where a TSP query is a TSP instance with $n$ points. In the literature, many advanced TSP solvers are developed to find high-quality solutions. Such solvers can solve some TSP instances efficiently but may take an extremely long time for some other instances. Due to the diversity of TSP instances, it is well-known that there exists no universal best solver dominating all other solvers on all possible TSP instances. To solve TSP efficiently, in addition to developing new TSP solvers, it needs to find a per-instance solver for each TSP instance, which is known as the TSP solver selection problem. In this paper, for the first time, we propose a deep learning framework, \CTAS, for TSP solver selection in an end-to-end manner. Specifically, \CTAS exploits deep convolutional neural networks to extract informative features from TSP instances and involves data argumentation strategies to handle the scarcity of labeled TSP instances. Moreover, to support large scale TSP solver selection, we construct a challenging TSP benchmark dataset with 6,000 instances, which is known as the largest TSP benchmark. Our \CTAS achieves over 2$\times$ speedup of the average running time, comparing the single best solver, and outperforms the state-of-the-art statistical models.

Multi-domain learning (MDL) refers to learning a set of models simultaneously, where each model is specialized to perform a task in a particular domain. Generally, a high labeling effort is required in MDL, as data needs to be labeled by human experts for every domain. Active learning (AL) can be utilized in MDL to reduce the labeling effort by only using the most informative data. The resultant paradigm is termed multi-domain active learning (MDAL). In this work, we provide an exhaustive literature review for MDAL on the relevant fields, including AL, cross-domain information sharing schemes, and cross-domain instance evaluation approaches. It is found that the few studies which have been directly conducted on MDAL cannot serve as off-the-shelf solutions on more general MDAL tasks. To fill this gap, we construct a pipeline of MDAL and present a comprehensive comparative study of thirty different algorithms, which are established by combining six representative MDL models and five commonly used AL strategies. We evaluate the algorithms on six datasets involving textual and visual classification tasks. In most cases, AL brings notable improvements to MDL, and the naive BvSB (best vs. second best) Uncertainty strategy can perform competitively with the state-of-the-art AL strategies. Besides, BvSB with the MAN (multinomial adversarial networks) model can consistently achieve top or above-average performance on all the datasets. Furthermore, we qualitatively analyze the behaviors of the well-performed strategies and models, shedding light on their superior performance in the comparison. Finally, we recommend using BvSB with the MAN model in the application of MDAL due to their good performance in the experiments., Comment: This work has been accepted by IEEE-TETCI

Over the past few years, various word-level textual attack approaches have been proposed to reveal the vulnerability of deep neural networks used in natural language processing. Typically, these approaches involve an important optimization step to determine which substitute to be used for each word in the original input. However, current research on this step is still rather limited, from the perspectives of both problem-understanding and problem-solving. In this paper, we address these issues by uncovering the theoretical properties of the problem and proposing an efficient local search algorithm (LS) to solve it. We establish the first provable approximation guarantee on solving the problem in general cases. Extensive experiments involving 5 NLP tasks, 8 datasets and 26 NLP models show that LS can largely reduce the number of queries usually by an order of magnitude to achieve high attack success rates. Further experiments show that the adversarial examples crafted by LS usually have higher quality, exhibit better transferability, and can bring more robustness improvement to victim models by adversarial training.

Single-phase AC-DC power factor correction (PFC) converters have instantaneous input and output power imbalances, resulting in second harmonic components in the output voltage. Usually the external voltage loop is designed as a low-bandwidth mode, that is, a low-pass filter (LPF) is introduced to eliminate the influence of the output voltage harmonics on the reference current, which reduces the dynamic performance of the system. Different from the existing LPF harmonic influence suppression scheme, this paper proposes a multi-dimensional harmonic current feedforward compensation control scheme, which extracts the odd harmonics in the input current through a band-pass filter, and converts the odd harmonics in the input current into even harmonics by dq coordinate transformation wave. And the converted even harmonics are fed forward to cancel the even harmonics in the voltage outer loop, thereby suppressing the harmonic distortion of the reference current. Taking a single-phase Boost PFC circuit as an example, an experimental prototype was built and the feasibility verification of the proposed scheme was completed. The results show that compared with the traditional LPF scheme, the proposed scheme improves its dynamic response by about 70% while ensuring high-quality input current.

Generalization, i.e., the ability of solving problem instances that are not available during the system design and development phase, is a critical goal for intelligent systems. A typical way to achieve good generalization is to learn a model from vast data. In the context of heuristic search, such a paradigm could be implemented as configuring the parameters of a parallel algorithm portfolio (PAP) based on a set of “training” problem instances, which is often referred to as PAP construction. However, compared to the traditional machine learning, PAP construction often suffers from the lack of training instances, and the obtained PAPs may fail to generalize well. This article proposes a novel competitive co-evolution scheme, named co-evolution of parameterized search (CEPS), as a remedy to this challenge. By co-evolving a configuration population and an instance population, CEPS is capable of obtaining generalizable PAPs with few training instances. The advantage of CEPS in improving generalization is analytically shown in this article. Two concrete algorithms, namely, CEPS-TSP and CEPS-VRPSPDTW, are presented for the traveling salesman problem (TSP) and the vehicle routing problem with simultaneous pickup–delivery and time windows (VRPSPDTW), respectively. The experimental results show that CEPS has led to better generalization, and even managed to find new best-known solutions for some instances.

The Vehicle Routing Problem with Simultaneous Pickup-Delivery and Time Windows (VRPSPDTW) has attracted much research interest in the last decade, due to its wide application in modern logistics. Since VRPSPDTW is NP-hard and exact methods are only applicable to small-scale instances, heuristics and meta-heuristics are commonly adopted. In this paper we propose a novel Memetic Algorithm with efficienT local search and Extended neighborhood, dubbed MATE, to solve this problem. Compared to existing algorithms, the advantages of MATE lie in two aspects. First, it is capable of more effectively exploring the search space, due to its novel initialization procedure, crossover and large-step-size operators. Second, it is also more efficient in local exploitation, due to its sophisticated constant-time-complexity move evaluation mechanism. Experimental results on public benchmarks show that MATE outperforms all the state-of-the-art algorithms, and notably, finds new best-known solutions on 12 instances (65 instances in total). Moreover, a comprehensive ablation study is also conducted to show the effectiveness of the novel components integrated in MATE. Finally, a new benchmark of large-scale instances, derived from a real-world application of the JD logistics, is introduced, which can serve as a new and more challenging test set for future research.

Exploiting parallelism is becoming more and more important in designing efficient solvers for computationally hard problems. However, manually building parallel solvers typically requires considerable domain knowledge and plenty of human effort. As an alternative, automatic construction of parallel portfolios (ACPP) aims at automatically building effective parallel portfolios based on a given problem instance set and a given rich configuration space. One promising way to solve the ACPP problem is to explicitly group the instances into different subsets and promote a component solver to handle each of them. This paper investigates solving ACPP from this perspective, and especially studies how to obtain a good instance grouping. The experimental results on two widely studied problem domains, the boolean satisfiability problems (SAT) and the traveling salesman problems (TSP), showed that the parallel portfolios constructed by the proposed method could achieve consistently superior performances to the ones constructed by the state-of-the-art ACPP methods, and could even rival sophisticated hand-designed parallel solvers.

In this article, we successfully apply the novel model selection method, Bayesian composite model space approach which has been used to map quantitative trait loci (QTL) for allelic substitution model, to map QTL for variance component model. The novel model selection approach has two advantages compared to the reversible jump Markov chain Monte Carlo method. Firstly, it mixes well due to the fixedness of the model dimension; secondly, it can map multiple QTL with higher power especially in genome-wide QTL mapping; finally, in the new method, it is also easy to incorporate our prior information about the variance components, which may bring precise estimate for variance components. A series of simulation experiments were conducted to demonstrate the general characters of the proposed method. The computer program is written in FORTRAN language, which is also built into a software "BayesMapQTL", and they also can be used for real data analysis and are available for request.

Using the data of crosses of multiple of inbred lines for mapping QTL can increase QTL detecting power compared with only cross of two inbred lines. Although many fixed-effect model methods have been proposed to analyze such data, they are largely based on one-QTL model or main effect model, and the interaction effects between QTL are always neglected. However, effectively separating the interaction effects from the residual error can increase the statistical power. In this article, we both extended the novel Bayesian model selection method and Bayesian shrinkage estimation approaches to multiple inbred line crosses. With two extensions, interacting QTL are effectively detected with high solution; in addition, the posterior variances for both main effects and interaction effects are also subjected to full Bayesian estimate, which is more optimal than two step approach involved in maximum-likelihood. A series of simulation experiments have been conducted to demonstrate the performance of the methods. The computer program written in FORTRAN language is freely available on request.

Cloud service composition problem (CSCP) is usually long-term based in practice. A logical request is to maximize end users' long-term benefit. Thus, the overall long-term QoS properties of the composite service should be optimized and the users' requirements during the period should be satisfied. However, the benefit-maximization has not been considered under the background of long-term based CSCP in existing research yet. To fill this gap, in this paper, a new formulation LCSCP is proposed to define the long-term based CSCP as an optimization problem. Then, for the sake of efficiency, three meta-heuristic approaches (i.e, Genetic Algorithm, Simulated Annealing and Tabu Search) are studied. Comprehensive experiments are designed and conducted to test their various aspects of performance on different test sets with different workflows. Experimental results provide a basic perspective of how these three widely adopted meta-heuristic frameworks work on this new problem, which can be baseline work for further research.

With the further research of sustainable development strategy in the world, more and more enterprises have been identifying the harmonious development between economy and ecological environment. Under such background, reverse logistics appears. As a new field in logistics management, reverse logistics has attracted more and more attention in the circle of business and theory. It has so many advantages as reducing production costs, increasing customers' degree of satisfaction and enhancing competitive ability of companies, which are of great significance to our business practices. This article discusses the meaning, role of reverse logistics and its development situation in foreign countries. Starting from analyzing the process and characteristics of reverse logistics under e-business environment, the paper makes a detailed analysis on the e-business model of reverse logistics and then puts forward three forms supporting the reverse logistics and suggestions to implement the reverse logistics. By using reverse logistics, enterprises can achieve the goals of sustainable development and help them to compete in their industries, especially when confronting intense competition and low profit margins. So it is urgent for enterprises to implement reverse logistics.

Amaranthus tricolor L. is a C4 plant, which is consumed as a major leafy vegetable in some tropical countries. Under conditions of high temperature and short daylight, Am. tricolor readily bolts and blooms, degrading leaf quality. A preliminary in vitro flowering study demonstrated that the flowering control pathway in Am. tricolor may differ from that of Arabidopsis. Nevertheless, no transcriptome analysis of the flowering process in Amaranthus has been conducted. To study Am. tricolor floral regulatory mechanisms, we conducted a large-scale transcriptome analysis--based on Illumina HiSeq sequencing of cDNA libraries generated from Am. tricolor at young seedling (YSS), adult seedling (ASS), flower bud (FBS), and flowering (FS) stages. A total of 99,312 unigenes were obtained. Using BLASTX, 43,088 unigenes (43.39%) were found to have significant similarity with accessions in Nr, Nt, and Swiss-Prot databases. Of these unigenes, 11,291 were mapped to 266 KEGG pathways. Further analysis of the four digital transcriptomes revealed that 735, 17,184, 274, and 206 unigenes were specifically expressed during YSS, ASS, FBS, and FS, respectively, with 59,517 unigenes expressed throughout the four stages. These unigenes were involved in many metabolic pathways related to in vitro flowering. Among these pathways, 259 unigenes were associated with ubiquitin-mediated proteolysis, indicating its importance for in vitro flowering in Am. tricolor. Other pathways, such as circadian rhythm and cell cycle, also had important roles. Finally, 26 unigenes were validated by qRT-PCR in samples from Am. tricolor at YSS, ASS, FBS, and FS; their differential expressions at the various stages indicate their possible roles in Am. tricolor growth and development, but the results were somewhat similar to Arabidopsis. Because unigenes involved in many metabolic pathways or of unknown function were revealed to regulate in vitro plantlet growth and flowering in Am. tricolor, the process appears to be highly complex in this species.