Most papers are machine-translated to English; some are originally in English.
Showing 35 of 35 papers in Integration Theory Of Computer Science
We point out that neural networks are not black boxes, and their generalization stems from the ability to dynamically map a dataset to the extrema of the model function. We further prove that the number of extrema in a neural network is positively correlated with the number of its parameters. We the…
[Objective] To address the "technical silo" issue in the full drug discovery and development pipeline, where target screening, molecular design, efficacy evaluation, and clinical translation rely on isolated models and lack a lightweight end-to-end integrated framework. [Methods] Based on the AutoGe…
We point out that neural networks are not black boxes, and their generalization stems from the ability to dynamically map a dataset to the local extrema of the model function. We further prove that the number of local extrema in a neural network is positively correlated with the number of its parame…
Bone age assessment is a commonly used method for detecting endocrine and growth abnormalities in children; however, low-quality hand X-ray images in deep learning methods reduce the final assessment accuracy. To address this issue, we propose an alignment network that increases the region of intere…
In recent years, the application of unmanned aerial vehicles (UAVs) has become increasingly widespread, and the coordination of multiple UAVs to accomplish tasks has substantially improved operational efficiency. Motivated by this phenomenon, numerous scholars have devoted themselves to research on …
To address the problems of low delivery timeliness and poor customer value in the logistics industry resulting from increasing vehicle ownership and escalating traffic congestion, this study proposes a time-varying routing optimization method for truck-drone collaborative delivery, comprehensively c…
To enhance the solution efficiency for the Multi-Depot Vehicle Routing Problem (MDVRP), we propose an end-to-end deep reinforcement learning framework. First, MDVRP is formulated as a Markov Decision Process (MDP), encompassing definitions of state, action, and reward. Additionally, we propose an im…
Text representation needs to address the ambiguity of textual terms and accurately define their semantic features within specific contextual environments. To tackle the problems of word polysemy and contextual characteristics, we propose an SCDVAB model for text semantic disambiguation. The main inn…
To address the issues of time-consuming training, large memory consumption, and difficulty in model updating associated with traditional offline hashing algorithms, as well as the phenomenon of substantial label loss in real-world image datasets, we propose an online hashing algorithm for balanced l…
To achieve accurate classification of breast pathology WSI images, we propose a gated convolutional neural network classification method based on hybrid connections. A hybrid module incorporating local residual connections and global dense connections is constructed, with squeeze-and-excitation gate…
Mobile crowdsensing data contains image and spatio-temporal contextual information that can be utilized for detecting changes in street view images; however, such data is typically low-quality and non-standard. To accurately detect changes in street scenes, this paper primarily addresses the data qu…
Link prediction is an important problem in the field of data mining. Similarity methods based on random walks generally assume that the probability of a walking particle transferring to neighboring nodes is equal, ignoring the influence of node degree values on the transfer probability. To address t…
Named Data Networking (NDN), as a novel Internet architecture, aims to address the ever-increasing data traffic. Based on its consumer-driven content retrieval model, NDN naturally supports consumer mobility. However, producer mobility remains a challenging problem, requiring additional mechanisms t…
Existing mobile crowdsensing task recommendation methods share common drawbacks: on one hand, they fail to fully consider the impact of spatio-temporal information on worker preferences, resulting in low recommendation accuracy; on the other hand, they ignore the influence of task popularity on reco…
In the IoT service discovery process, users typically express their requirements through their own intentions, whereas service descriptions constitute explanations of service functionalities; consequently, mismatches between these two elements impact the accuracy of service discovery. Simultaneously…
In deploying blockchain-based IoT applications, technical obstacles have emerged, including lack of fine-grained privacy protection, low transaction processing efficiency, high latency, and insufficient flexibility and dynamism. To further advance the popularization and practical deployment of block…
To further improve the accuracy and speed of optical flow matching in visual SLAM, this paper proposes an improved optical flow matching algorithm that integrates Inertial Measurement Unit (IMU) information to remove motion blur. The algorithm first employs a point spread function computed from IMU …
SM9-IBS is an identity-based signature algorithm industry standard published by China in 2016. Although identity-based signature algorithms reduce the complexity of system management of user public keys, they suffer from the difficult problem of key revocation. Furthermore, the special structure of …
To address the issues of the TextRank algorithm ignoring syntactic information and thematic information when extracting document keywords, we propose a document keyword extraction model based on syntactic analysis and thematic distribution. The model extracts document keywords through a two-stage pr…
Currently, research on the Stochastic Duration Distributed Resource-Constrained Multi-Project Scheduling Problem (SDRCMPSP) is scarce, with most studies focusing on static scheduling schemes that cannot adjust and optimize strategies in real-time in response to environmental changes or promptly addr…
With the goal of enhancing the cognitive reasoning capability of recommendation algorithm models and overcoming the current limitation where traditional recommendation algorithms heavily rely on data quality, resulting in constrained performance, we propose an implicit deep collaborative recommendat…
To reduce data storage requirements and transmission overhead in wireless Internet of Things (IoT), a local directed acyclic graph blockchain scheme (LDB) based on directed acyclic graph (DAG) is proposed. This scheme addresses node storage constraints and transmission overhead by enabling wireless …
We propose a multi-description coding image enhancement method based on joint side and central decoding feature learning, which simultaneously considers both side decoding image enhancement and central decoding image enhancement, thereby enabling better network training through joint optimization of…
Aspect-level sentiment analysis aims to determine the sentiment polarity toward specific aspects in reviews; however, few studies have investigated the impact of complex sentences on sentiment classification. Based on this, we propose an aspect-level sentiment analysis model based on BERT and a self…
To address the limitations of conventional salient object detection methods in detecting multiple salient objects at different scales, we propose a salient object detection algorithm based on deep reuse of multi-scale features. The network model consists of vertically stacked bidirectional dense fea…
Deep Reinforcement Learning (DRL) can be widely applied in the field of urban traffic signal control; however, in existing research, the vast majority of DRL agents utilize only the current traffic state for decision-making, resulting in limited control effectiveness under conditions of significant …
In mobile crowdsensing systems, tasks exhibit spatiotemporal coverage overlap, which may lead to redundant data collection and consequently cause data redundancy issues. To address this, we propose a task allocation method that can simultaneously control both intra-task and inter-task data redundanc…
To address the challenges of next-generation networks in coverage, deployment cost, and capacity, Mobile Edge Computing (MEC) often necessitates the assistance of relay nodes to execute computation-intensive and delay-sensitive tasks. This paper first introduces the fundamental architecture of relay…
Max-SAT is the optimization version of the SAT problem, with the objective of finding a variable assignment that maximizes the number of satisfied clauses in a given clause set. This problem is typically NP-hard. With the in-depth development of big data and artificial intelligence, existing algorit…
Given that edge clouds lack more powerful computational processing capabilities compared to central clouds, they are susceptible to unnecessary scaling jitter or insufficient resource processing capability when handling dynamic workloads. This paper therefore conducts experimental evaluations on mic…
When swarm intelligence heuristic algorithms solve the Discounted {0-1} Knapsack Problem (D{0-1}KP), some repair and optimization strategy is required to transform abnormal encoding individuals into encoding individuals that satisfy the solution constraints, in order to improve solution efficiency a…
Recommendation systems have found extensive applications across diverse domains, profoundly influencing daily life. Training an effective recommendation system typically requires vast amounts of "user-item" interaction data; however, data obtained in practice is often extremely sparse, frequently le…
To address the issue of attackers exploiting background knowledge and other information to launch attacks during user location privacy protection, this paper proposes a location privacy protection method for mobile terminals. The scheme leverages k-anonymity and local differential privacy techniques…
Graph Convolutional Networks (GCNs) can extract effective information from graph data through graph convolutions, but they are vulnerable to adversarial attacks that degrade model performance. While adversarial training can enhance neural network robustness, the discrete nature of graph structures a…
To address issues such as poor car-following stability, ineffective tracking, or unsafe conditions caused by significant speed fluctuations of vehicles in congested environments, a multi-objective optimization car-following scheme based on vehicle models and deep reinforcement learning is proposed. …
