With the development of the Internet of Things (IoT), Federated Learning (FL) is extensively employed in smart cities and industrial IoT, involving numerous heterogeneous devices with varying computational and storage capabilities. Traditional FL ...

This paper proposes an adaptive federated learning framework for heterogeneous devices (AdaptFL). AdaptFL first automatically allocates a model suitable for its resources to each client through the Device Resource-aware Neural ...

• AdaptFL tackles device heterogeneity in FL while preserving performance and privacy.
• DRNAS enables evaluation of searched models without relying on public datasets.
• ESKD enables heterogeneous model distribution and aggregation in ...

Energy efficiency of Convolutional Neural Networks (CNNs) has become an important area of research, with various strategies being developed to minimize the power consumption of these models. Previous efforts, including techniques like model ...

Display Omitted

• Introduced the concept of complementarity in Neural Networks.
• Proposed an on-device AI using the synergy of two small, complementary CNNs.
• A confidence-based mechanism selects accurate predictions from two small CNNs dynamically.

With the rapid development of the Industrial Internet of Things (IIoT), ensuring the security and privacy of network traffic has become particularly important. Classifying and identifying encrypted traffic is a critical step in enhancing network ...

Display Omitted

• Proposed an innovative serial network architecture.
• Developed an advanced pre-trained model specifically designed for encrypted traffic analysis.
• Significantly improved the speed and accuracy of encrypted traffic classification in ...

Edge applications are increasingly empowered by deep neural networks (DNN) and face the challenges of adapting or retraining models for the changes in input data domains and learning tasks. The existing techniques to enable DNN retraining on edge ...

• Propose a lightweight in-memory hyperparameter tuning method for retraining.
• Improve deployment efficiency by implementing an online resource profiler.
• Achieve continuous training accuracy improvement by using a rule engine.
• ...

Training Deep Learning (DL) models are becoming more time-consuming, thus interruptions to the training processes are inevitable. We can obtain an optimal checkpointing interval to minimize the fault tolerance overhead for a HPC (High Performance ...

• A novel checkpointing interval problem for exploratory DL training jobs based on model convergence progress.
• An approach to compute optimal check-pointing configuration for a DL training job, minimizing the fault-tolerant overhead.

Federated learning has emerged as a prominent solution for the collaborative training of machine learning models without exchanging local data. However, existing approaches often impose rigid constraints on model heterogeneity, limiting the ...

• Focusing on data generation in federation rather than training task-specific models.
• Eliminating the requirement for a shared objective in model training.
• Safeguarding the privacy of local data and models, as well as generated.

As an emerging distributed machine learning paradigm, federated learning has been extensively used in the domain of cloud–edge computing to collaboratively train models without uploading their raw data. However, the existing federated learning ...

• Introduces FedCD, a novel framework for personalized federated learning.
• Implements a local training strategy leveraging cross-client knowledge.
• Designs a global weighted aggregation mechanism based on cross-client importance.
• ...

In recent years, there has been a significant increase in the processing of massive amounts of data, driven by the growing demands of mobile systems, parallel and distributed architectures, and real-time systems. This applies to various types of ...

• Main features of an efficient address predictor are outlined.
• New predictor (SVM4AP) that improves the performance-cost trade-off is proposed and evaluated.
• Demonstrated the suitability of SVM4AP for memory-constrained embedded ...

In federated learning, data heterogeneity leads to significant differences in the local models learned by the clients, thereby affecting the performance of the global model. To address this issue, contrast federated learning algorithms increase ...

• Sliding-window model contrastive learning introduces more influential historical local models as positive or negative samples, enabling local models to learn more robust and generalizable representations.
• Due to differences in the ...

Federated learning (FL) is an innovative distributed learning paradigm that permits multiple parties to train models collaboratively while protecting individual privacy. However, it encounters security challenges, making it vulnerable to several ...

• Analyze the impact of label-flipping attacks on diverse adversarial settings.
• We propose a server-side defense mechanism to counter the poisoning attacks.
• We implement temporal attacks (middle and end rounds) and assess defense ...

Genomes hold the complete genetic information of an organism. Examining and analyzing genomic data plays a critical role in properly understanding an organism, particularly the main characteristics, functionalities, and evolving nature of harmful ...

• A framework is developed, called F4GDA, for the analysis of viral genomic data in various forms.
• F4GDA can interpret and analyze genomes with a focus on both nucleotide and amino acid sequence variations.
• As a case study, the F4GDA ...

Protein representation learning plays a crucial role in obtaining a comprehensive understanding of biological regulatory mechanisms and in developing proteins and drugs for therapeutic purposes. However, labeled proteins, such as sequenced and ...

• Pro-CoRL integrates LLMs & PLMs for enhanced multi-knowledge contrastive learning.
• Addresses data sparsity, extracts hidden features across multiple protein knowledge.
• Optimizes joint objective to overcome training challenges from ...

Motivated by the large amount of available satellite data and increasing interest in the study of urbanization, this paper presents a way for better supervision of urbanization, as more and more people are looking to increase their quality of ...

• A multi-spectral satellite imagery cloud service.
• Computation of spectral indices that highlight patterns in satellite images.
• Reliable and easy-to-use tool accurately mapping the build-up in cities.
• Study of the urban sprawl ...

Pathological diagnosis is considered the gold standard in cancer diagnosis, playing a crucial role in guiding treatment decisions and prognosis assessment for patients. However, achieving accurate diagnosis of pathology images poses several ...

• The adoption of digital pathology facilitates the application of deep learning in pathological image classification.
• A brief overview of task-specific models and foundation models for pathological images classification is presented.

The ADMM (Alternating Direction Method of Multipliers) optimization framework is known for its property of decomposition and assembly, which effectively bridges distributed computing and optimization algorithms, making it well-suited for ...

• FIADMM uses inertial acceleration and adaptive iterations to speed up convergence.
• Fast linear convergence rate O ( 1 / k ) proven.
• Validated on six datasets for SVR and Probit models, showing good gains.

Satellite communication and Low Earth Orbit (LEO) satellites are important components of the 6G network, widely used for Earth observation tasks due to their low cost and short return period, making them a key technology for 6G network ...

• In the existing satellite federated learning, the efficiency of the machine learning process is often compromised due to limited communication links, resulting in high communication costs. Moreover, machine learning models increase in size ...

Wind speed time series forecasting has been widely used in wind power generation. However, the nonlinear and non-stationary characteristics of wind speed make accurate wind speed forecasting a difficult task. In recent years, the rapid ...

• This research proposes an optimized hybrid wind speed forecasting model.
• FIG is adopted as a preprocessing operation, which better preserves the trend characteristics.
• Swarm intelligence algorithm is designed to search for the ...

Anomaly detection in edge–cloud scenarios stands as a critical means to ensure the security of network environment. Federated learning (FL)-based anomaly detection combines multiple data sources and ensures data privacy, making it a promising ...

• An iterative federated clustering ensemble algorithm (IFCEA) for edge–cloud anomaly detection environments.
• Client selection method and aggregation weights designed for the FL-based anomaly detection system.
• A cluster ...

Computational drug repositioning is a vital path to improve efficiency of drug discovery, which aims to find potential Drug–Disease Associations (DDAs) to develop new effects of the existing drugs. Many approaches detected novel DDAs ...

• Gaussian similarity kernel fusion algorithm measures similarity of drug and disease.
• DDA network is induced from convolutional graph embeddings of drugs and diseases.
• Collaborative learning model offers the comprehensive inferences ...

Traffic classification is essential for network management and optimization, enhancing user experience, network performance, and security. However, evolving technologies and complex network environments pose challenges. Recently, researchers have ...

• A federated learning framework that combines split learning with mutual learning.
• Applicable for traffic classification and compatible with model heterogeneity.
• Reduction of computational and storage overhead on the client side.