Zero-sample diagnostic methods have gained recognition in addressing the scarcity of gearbox fault samples, thereby being regarded as a promising technique to guarantee gearbox safety. However, historical zero-sample approaches typically neglect the use of multimodal noncontact sensing data and rarely consider the interpretability of the diagnostic process. This oversight limits their application in industrial environments that require high reliability or operate under extreme conditions. Therefore, this article presents a composite neuro-fuzzy system-guided cross-modal zero-sample diagnostic framework, termed FCZD-IA, which employs infrared thermography and acoustic data to monitor gearbox conditions. Specifically, FCZD-IA uses a proposed composite neural system as a decision-maker in the diagnostic task, while integrating a deep backbone network to discriminatively learn high-level fault features from multimodal data. Moreover, a specific training strategy is designed to guide the learning process of the FCZD-IA to promote robust and interpretable zero-sample diagnostics. Comprehensive experimental results validate the effectiveness of the proposed framework and its superiority over other competitive methods.

Online advertising has been one of the most important sources for industry's growth, where the demand-side platforms (DSP) play an important role via bidding to the ad exchanges on behalf of their advertiser clients. Since more and more ad exchanges have shifted from second to first price auctions, it is challenging for DSPs to adjust bidding strategy in the volatile environment. Recent studies on bid shading in first-price auctions may have limited performance due to relatively strong hypotheses about winning probability distribution. Moreover, these studies do not consider the incentive of advertiser clients, which can be crucial for a reliable advertising platform. In this work, we consider both the optimization of bid shading technique and the design of internal auction which is ex-post incentive compatible (IC) for the management of a DSP. Firstly, we prove that the joint design of bid shading and ex-post IC auction can be reduced to choosing one monotone bid function for each advertiser without loss of optimality. Then we propose a parameterized neural network to implement the monotone bid functions. With well-designed surrogate loss, the objective can be optimized in an end-to-end manner. Finally, our experimental results demonstrate the effectiveness and superiority of our algorithm.

Convolutional neural network (CNN)-based fault detection approaches based on multisource signals have attracted increasing interest from the research community and industrial practices, thanks to the powerful feature representation capability of CNN and the rapid development of sensor technology. Various strategies have been applied in existing CNN-based diagnostic models to learn features from 1-D real-valued multivariate data. However, the distribution gap and the intrinsic correlations among multisource mechanical signals during the learning process have been rarely considered, which may lead to suboptimal fault identification results. To tackle this issue, this article proposes a cross-modal fusion convolutional neural network (CMFCNN) for mechanical fault diagnosis, which performs modality-specific and cross-modal feature representation on multisource data. Specifically, CMFCNN adopts two parallel modality-specific networks and a cross-modal knowledge-sharing network to fully explore independent and shared features from the multisource mechanical signals. To achieve effective feature propagation and fusion, a cross-modal fusion module is introduced to integrate cross-modal features and pass the fused information to the next layer. Moreover, to alleviate overfitting and achieve a better diagnostic performance of the framework, an online soft-label training algorithm is adopted in the CMFCNN training phase. Extensive experimental results on the cylindrical rolling bearing dataset and the planetary gearbox dataset validate that the proposed CMFCNN outperforms seven state-of-the-art methods significantly, especially under strong noise conditions.

Image manipulation detection (IMD) is of vital importance as faking images and spreading misinformation can be malicious and harm our daily life. IMD is the core technique to solve these issues and poses challenges in two main aspects: (1) Data Uncertainty, i.e., the manipulated artifacts are often hard for humans to discern and lead to noisy labels, which may disturb model training; (2) Model Uncertainty, i.e., the same object may hold different categories (tampered or not) due to manipulation operations, which could potentially confuse the model training and result in unreliable outcomes. Previous works mainly focus on solving the model uncertainty issue by designing meticulous features and networks, however, the data uncertainty problem is rarely considered. In this paper, we address both problems by introducing an uncertainty-guided learning framework, which measures data and model uncertainties by a novel Uncertainty Estimation Network (UEN). UEN is trained under dynamic supervision, and outputs estimated uncertainty maps to refine manipulation detection results, which significantly alleviates the learning difficulties. To our knowledge, this is the first work to embed uncertainty modeling into IMD. Extensive experiments on various datasets demonstrate state-of-the-art performance, validating the effectiveness and generalizability of our method.

Unmanned-aerial-vehicle (UAV)-assisted communications has attracted increasing attention recently. This article investigates air–ground coordinated communications system, in which trajectories of air UAV base stations (UAV-BSs) and access control of ground users (GUs) are jointly optimized. We formulated this optimization problem as a mixed cooperative–competitive game, where each GU competes for the limited resources of UAV-BSs to maximize its own throughput by accessing a suitable UAV-BS, and UAV-BSs cooperate with each other and design their trajectories to maximize the definedfair throughputto improve the total throughput and keep the GU fairness. Moreover, the action space of GUs is discrete, while that of UAV-BS is continuous. To tackle this hybrid action space issue, we transform the discrete actions into continuous action probabilities and propose a multiagent deep reinforcement learning (MADRL) approach, named air–ground probabilistic multiagent deep deterministic policy gradient (AG-PMADDPG). With well-designed rewards, AG-PMADDPG can coordinate two types of agents, UAV-BSs and GUs, to achieve their own objectives based on local observations. Simulation results demonstrate that AG-PMADDPG can outperform the benchmark algorithms in terms of throughput and fairness.

Online platforms sell advertisements via auctions (e.g., VCG and GSP auction) and revenue maximization is one of the most important tasks for them. Many revenue increment methods are proposed, like reserve pricing, boosting, coupons and so on. The novelty of coupons rests on the fact that coupons are optional for advertisers while the others are compulsory. Recent studies on coupons have limited applications in advertising systems because they only focus on second price auctions and do not consider the combination with other methods. In this work, we study the coupon design problem for revenue maximization in the widely used VCG auction. Firstly, we examine the bidder strategies in the VCG auction with coupons. Secondly, we cast the coupon design problem into a learning framework and propose corresponding algorithms using the properties of VCG auction. Then we further study how to combine coupons with reserve pricing in our framework. Finally, extensive experiments are conducted to demonstrate the effectiveness of our algorithms based on both synthetic data and industrial data.

Traditionally, the model time in agent-based simulations is advanced in fixed time steps. However, a purely time-stepped execution is inefficient in situations where the states of individual agents are independent of other agents and thus easily predictable far into the simulated future. In this work, we propose a method to accelerate microscopic traffic simulations based on identifying independence among agent state updates. Instead of iteratively updating an agent's state throughout a sequence of time steps, a computationally inexpensive "fast-forward" function advances the agent's state to the time of its earliest possible interaction with other agents. To demonstrate the approach in practice, we present an algorithm to efficiently determine intervals of independence in microscopic traffic simulations and derive a fast-forward function for the popular Intelligent Driver Model (IDM). In contrast to existing acceleration approaches based on reducing the level of model detail, our approach retains the microscopic nature of the simulation. A performance evaluation is performed in a synthetic scenario and on the road network of the city of Singapore. At low traffic densities, we achieved a speedup of up to 2.8, whereas at the highest considered densities, only few opportunities for fast-forwarding could be identified. The algorithm parameters can be tuned to control the overhead of the approach.

A common approach of parallelising an agent-based road traffic simulation is to partition the road network into sub-regions and assign computations for each subregion to a logical process (LP). Inter-process communication for synchronisation between the LPs is one of the major factors that affect the performance of parallel agent-based road traffic simulation in a distributed memory environment. Synchronisation overhead, i.e., the number of messages and the communication data volume exchanged between LPs, is heavily dependent on the employed road network partitioning algorithm. In this paper, we propose Neighbour-Restricting Graph-Growing (NRGG), a partitioning algorithm which tries to reduce the required communication between LPs by minimising the number of neighbouring partitions. Based on a road traffic simulation of the city of Singapore, we show that our method not only outperforms graph partitioning methods such as METIS and Buffoon, for the synchronisation protocol used, but also is more resilient than stripe spatial partitioning when partitions are cut more ?nely.

Road traffic simulation is a useful tool for studying road traffic and evaluating solutions to traffic problems. Large-scale agent-based road traffic simulation is computationally intensive, which triggers the need for conducting parallel simulation. This paper deals with the synchronization problem in parallel agent-based road traffic simulation to reduce the overall simulation execution time. We aim to reduce synchronization operations by introducing some redundant computation to the simulation. There is a trade-off between the benefit of reduced synchronization operations and the overhead of redundant computation. The challenge is to minimize the total overhead of redundant computation and synchronization. First, to determine the amount of redundant computation, we proposed a way to define extended layers of partitions in the road network. The sizes of extended layers are determined by the behavior of agents and the topology of road networks. Second, due to the dynamic nature of road traffic, a heuristic was proposed to adjust the amount of redundant computation according to traffic conditions during simulation run-time to minimize the overall simulation execution time. The efficiency of the proposed method was investigated in a parallel agent-based road traffic simulator using real-world network and trip data. Results have shown that the method can reduce synchronization overhead and improve the overall performance of the parallel simulation significantly.

Large-scale agent-based traffic simulation is a promising tool to study the road traffic and help solving traffic problems, such as congestion and high emission in megacities. Such simulation requires high computational resource which triggers the need for parallel computing. The parallelization of agent-based traffic simulations is generally performed by decomposing the simulation space into spatial subregions. The agent models contained by each subregion are executed by Logical Processes (LPs). As the simulated system evolves over the simulation time in individual LPs, synchronization among LPs is required due to data dependencies. Existing work has used global barriers for synchronization which is a type of synchronous synchronization method. However, global barriers have very low efficiency due to the waiting of processes at barriers. High synchronization overhead is still one of the major performance issues in parallel large-scale agent-based traffic simulations. In this paper, we proposed a novel asynchronous conservative synchronization strategy named Mutual Appointment (MA) to address this issue. MA removes global barriers and allows LPs to communicate individually. Since the efficiency of conservative synchronization relies on the lookahead of the simulated system, a heuristic was developed to increase the lookahead in agent-based traffic simulations. It takes advantage of the intrinsic uncertainties in traffic simulations. MA together with the lookahead heuristic forms the Relaxed Mutual Appointment (RMA) strategy. Its efficiency was investigated in the parallel agent-based traffic simulator SEMSim Traffic using real world traffic data. Experiment results showed that the MA strategy improved the speed-up of the parallel simulation compared to the barrier method, and the RMA strategy further improved the MA strategy by reducing the number of synchronization messages significantly.