Generating CT images from radiographs has immense clinical potential, offering a novel approach to low-cost and low-radiation medical imaging. This method can significantly ease the workload of radiologists. Current machine and deep learning approaches mainly utilize fully supervised learning methods for CT image generation tasks. However, the feature representations learned in fully supervised settings are typically task-specific, dependent on large quantities of labeled data, and have limited generalizability. Addressing these challenges, we introduce a contrastive learning-based method for CT image generation, which includes two phases: pre-training and fine-tuning. The pre-training phase aims to learn feature representations from unlabeled radiographs. Specifically, we devised two simple yet efficient radiograph data augmentation methods, converting the original data into two related but different views. These are then input into an encoder module to learn discriminative feature representations. Labeled data is used to learn the medical image generation task during the fine-tuning phase. In evaluations across the LIDC-IDRI lung CT and IXI brain MRI datasets, our CLCT-GAN model exhibits not only outstanding performance in lung CT reconstructions but also showcases remarkable adaptability to brain MRI data from IXI, surpassing previous state-of-the-art models in these diverse medical imaging benchmarks.

Graph-based Aspect-based Sentiment Classification (ABSC) approaches have yielded stateof-the-art results, expecially when equipped with contextual word embedding from pretraining language models (PLMs).However, they ignore sequential features of the context and have not yet made the best of PLMs.In this paper, we propose a novel model, BERT4GCN, which integrates the grammatical sequential features from the PLM of BERT, and the syntactic knowledge from dependency graphs.BERT4GCN utilizes outputs from intermediate layers of BERT and positional information between words to augment GCN (Graph Convolutional Network) to better encode the dependency graphs for the downstream classification.Experimental results demonstrate that the proposed BERT4GCN outperforms all state-of-the-art baselines, justifying that augmenting GCN with the grammatical features from intermediate layers of BERT can significantly empower ABSC models.

Abstract Dropout has been proven to be an effective technique for regularizing and preventing the co-adaptation of neurons in deep neural networks (DNN). It randomly drops units with a probability of p during the training stage of DNN to avoid overfitting. The working mechanism of dropout can be interpreted as approximately and exponentially combining many different neural network architectures efficiently, leading to a powerful ensemble. In this work, we propose a novel diversification strategy for dropout, which aims at generating more different neural network architectures in less numbers of iterations. The dropped units in the last forward propagation will be marked. Then the selected units for dropping in the current forward propagation will be retained if they have been marked in the last forward propagation, i.e., we only mark the units from the last forward propagation. We call this new regularization scheme Tabu dropout, whose significance lies in that it does not have extra parameters compared with the standard dropout strategy and is computationally efficient as well. Experiments conducted on four public datasets show that Tabu dropout improves the performance of the standard dropout, yielding better generalization capability.

SGD (Stochastic Gradient Descent) is a popular algorithm for large scale optimization problems due to its low iterative cost. However, SGD can not achieve linear convergence rate as FGD (Full Gradient Descent) because of the inherent gradient variance. To attack the problem, mini-batch SGD was proposed to get a trade-off in terms of convergence rate and iteration cost. In this paper, a general CVI (ConvergenceVariance Inequality) equation is presented to state formally the interaction of convergence rate and gradient variance. Then a novel algorithm named SSAG (Stochastic Stratified Average Gradient) is introduced to reduce gradient variance based on two techniques, stratified sampling and averaging over iterations that is a key idea in SAG (Stochastic Average Gradient). Furthermore, SSAG can achieve linear convergence rate of O((1 - μ/8C L)k) at smaller storage and iterative costs, where C ≥ 2 is the category number of training data. This convergence rate depends mainly on the variance between classes, but not on the variance within the classes. In the case of C ≪ N (N is the training data size), SSAG's convergence rate is much better than SAG's convergence rate of O((1 - μ/8N L)k). Our experimental results show SSAG outperforms SAG and many other algorithms.

Agent-based computing addresses the challenges in managing distributed computing systems and networks through monitoring, communication, consensus-based decision-making and coordinated actuation.As a result, intelligent agents and multi-agent systems have demonstrated the capability to use intelligence, knowledge representation and reasoning, and other social metaphors like 'trust', 'game' and 'institution', not only to address real-world problems in a human-like way but also to transcend human performance.This has had a transformative impact in many application domains, particularly in e-commerce, and also in planning, logistics, manufacturing, robotics, decision support, transportation, entertainment, emergency relief & disaster management, and data mining & analytics.As one of the largest and still growing research fields of Computer Science, agent-based computing today remains a unique enabler of inter-, multi-and trans-disciplinary research.The International Conference on Conference on Principles and Practice of Multi-Agent Systems (PRIMA) originally started in 1998 as a regional (Asia-Pacific) workshop and in the last decade it grew to become one of the leading and influential scientific conferences for research on multi-agent systems.Each year, PRIMA brings together active researchers, developers and practitioners from both academia and industry to showcase, share and promote research in several domains, ranging from foundations of agent theory and engineering aspects of agent systems, to emerging interdisciplinary areas of agent-based research.Previous successful editions were held in Nagoya, Japan (2009), Kolkata, India (2010), Wollongong, Australia (2011), Kuching, Malaysia (2012), Dunedin, New Zealand (2014), and Gold Coast Australia (2014).The last two editions were held in Phuket, Thailand (2016) and in Nice, France (2017).This issue contains selected papers from the eighteenth edition of PRIMA, which took place from 26 to 30 October 2015 in Bertinoro, FC, Italy.The conference received 94 submissions from 30 countries.From the 29 papers presented at the conference, the best theoretically-oriented ones were invited to submit an extended version for this special issue with Fundamenta Informaticae, while another special issue with the Journal of Agent-Oriented Software Engineering has been organized around the more practically-oriented papers.The contributions that were eventually submitted underwent a thorough two-or three-stage reviewing procedure, resulting in the ten papers in the present collection.They constitute the most recent advances in the theory and practice of multi-agent systems, with interesting intersections with other domains.

As a data structure for representing and manipulating Boolean functions, BDDs (Binary Decision Diagrams) are commonly used in many fields such as model-checking, system verification and so on. For saving space and improving operation speed, all the existing packages limit the number of variables to 216. However, such a limitation also restrains its applicability. In this paper, we present TiniBDD, an efficient implementation of a 232BDD package incorporating sub-allocation of memory and lightweight Garbage Collection as well as a new operator named as Satisfiable Assignment Operator. Compared with the well-known CUDD which is one of the best 216BDD packages that can be attained publicly, the experiments show TiniBDD has comparable performance.

A number of algorithms have been proposed for the Minimum Vertex Cover problem. However, they are far from satisfactory, especially on hard instances. In this paper, we introduce Edge Weighting Local Search (EWLS), a new local search algorithm for the Minimum Vertex Cover problem. EWLS is based on the idea of extending a partial vertex cover into a vertex cover. A key point of EWLS is to find a vertex set that provides a tight upper bound on the size of the minimum vertex cover. To this purpose, EWLS employs an iterated local search procedure, using an edge weighting scheme which updates edge weights when stuck in local optima. Moreover, some sophisticated search strategies have been taken to improve the quality of local optima. Experimental results on the broadly used DIMACS benchmark show that EWLS is competitive with the current best heuristic algorithms, and outperforms them on hard instances. Furthermore, on a suite of difficult benchmarks, EWLS delivers the best results and sets a new record on the largest instance.

Web service protocols are designed in XML formats so the message structures within are quite different from the conventional protocols. Therefore, the traditional formal verification techniques which have gain substantial achievements in practice, cannot be applied directly to them because their underlying models are written in Alice\Bob-style descriptions using high-level message formats instead of XML tags. In this paper, we propose a justification-oriented and automatic formal approach to verify, in the standard Dolev-Yao model, security properties expressed as epistemic notions for a Web service protocol, based on a fault-preserving mapping tool called SuD (SOAP under Dolev-Yao). Our approach can shed more light on Web service protocols in another perspective because the concerned properties to be verified are some inherent features of protocols.

Bounded model checking (BMC) has played an important role in verification of software, embedded systems and protocols. The idea of BMC is to encode finite state machine (FSM) and linear temporal logic (LTL) verification specification into satisfiability (SAT) instances, and then to search for a counterexample via various SAT tools. Improving encoding technology of BMC can generate a SAT instance easy to solve, and therefore is essential to improve the efficiency of BMC. In this paper, we improve the encoding of BMC by combining the characteristic of FSM state transition and semantics of LTL, get a simple and efficient recursion formula which is useful to efficiently generate SAT instances. We present an efficient algorithm to encode the modal operator (safety formula) in BMC. The experiments for comparative analysis shows that this encoding algorithm is more powerful than the existing two mainstream encoding algorithms in both the scale of generated SAT instances and the solving efficiency. The methodology presented in this paper is also valuable for optimization of other modal operator encodings in BMC.

The standard Kripke semantics of epistemic logics has been applied successfully to reasoning communication protocols under the assumption that the network is not hostile. This paper introduces a natural semantics of Kripke semantics called knowledge structure and, by this kind of Kripke semantics, analyzes communication protocols over hostile networks, especially on authentication protocols. Compared with BAN-like logics, the method is automatically implementable because it operates on the actual definitions of the protocols, not on some difficult-to-establish justifications of them. What is more, the corresponding tool called SPV (Security Protocol Verifier) has been developed. Another salient point of this approach is that it is justification-oriented instead of falsificationoriented, i.e. finding bugs in protocols.