Balanced Network Research Articles

Intrinsic dynamics of randomly clustered networks generate place fields and preplay of novel environments.

During both sleep and awake immobility, hippocampal place cells reactivate time-compressed versions of sequences representing recently experienced trajectories in a phenomenon known as replay. Intriguingly, spontaneous sequences can also correspond to forthcoming trajectories in novel environments experienced later, in a phenomenon known as preplay. Here, we present a model showing that sequences of spikes correlated with the place fields underlying spatial trajectories in both previously experienced and future novel environments can arise spontaneously in neural circuits with random, clustered connectivity rather than pre-configured spatial maps. Moreover, the realistic place fields themselves arise in the circuit from minimal, landmark-based inputs. We find that preplay quality depends on the network's balance of cluster isolation and overlap, with optimal preplay occurring in small-world regimes of high clustering yet short path lengths. We validate the results of our model by applying the same place field and preplay analyses to previously published rat hippocampal place cell data. Our results show that clustered recurrent connectivity can generate spontaneous preplay and immediate replay of novel environments. These findings support a framework whereby novel sensory experiences become associated with preexisting "pluripotent" internal neural activity patterns.

BENN: Balanced Ensemble Neural Network for Handling Class Imbalance in Big Data

ABSTRACTClass imbalance is a critical challenge in big data analytics, often leading to biased predictive models. This imbalance can lead to biased models that perform well on the majority class but poorly on the minority class. Many machine learning models tend to be biased towards the majority class because they aim to minimise overall error, often leading to poor performance on the minority class. This paper presents the balanced ensemble neural network, a novel solution to effectively address class imbalance in big data. Balanced ensemble neural network combines the robust capabilities of neural networks with the power of ensemble learning, incorporating class balancing strategies to ensure fair representation of minority classes. The methodology involves integrating multiple neural networks, each trained on balanced subsets of data using techniques like Synthetic Minority Over‐sampling Technique and Random Undersampling. This integration aims to leverage the strengths of individual networks while reducing their inherent biases. Our extensive experiments across various datasets reveal that BENN achieves an AUC‐ROC score of 0.94, surpassing other models such as random forest (0.88), support vector (0.84) and single neural net (0.80). It was also observed that BENN's performance is better compared to traditional neural network models and standard ensemble methods in key metrics like accuracy, precision, recall, F1‐score and AUC‐ROC. The results specifically highlight BENN's effectiveness in accurately classifying instances of minority classes, a notable challenge in many existing models. These findings underscore BENN's potential as a substantial advancement in handling class imbalance within big data environments, offering a promising direction for future research and application in machine learning.

Bonding versus bridging: disentangling effects of CEO social capital on firm exploratory innovation

CEOs and the influence of their social networks have been considered as important drivers for firm innovation. Yet, both theories and empirical evidence are lacking. To advance this stream of research, our paper fills the gap by exploring how bonding and bridging social capital influence exploratory innovation in firms. Through an examination of executive social ties and patent‐based technology indicators, our findings reveal nuanced relationships: Excessive bonding social capital exponentially hampers exploratory innovation due to information overload, while bridging social capital enhances exploratory innovation and mitigates over‐bonding effects (inverted U‐shape). We test our hypotheses with a sample of 7,540 firm‐year observations of public U.S. firms from 2000 to 2014. Our study not only contributes to research on strategic leadership, social capital, interfirm networks, and technological innovation but also offers important managerial implications. It suggests that CEOs and R&D directors foster balanced social networks in their pursuit of enhanced innovation management.

Gauss Markov and Flow Balanced Vector Radial Learning network traffic classification on IoT with SDN.

Recent evolution in connected devices modelled a massive stipulation for network traffic resources and classification. Software-defined networking (SDN) enables ML techniques with the Internet of Things (IoT) to automate network traffic. This helps to reduce accuracy and improves latency. Problems by conventional techniques to categorize network traffic acquired from IoT and assign resources can be resolved through SDN solutions. This manuscript proposes a proposed network traffic classification technique on IoT with SDN called Gauss Markov and Flow-balanced Vector Radial Learning (GM-FVRL). With the network traffic features acquired from the IoT devices, SDN-enabled Gauss Markov Correlation-based IoT Network Traffic Feature Extraction is applied to extort relevant network aspects. Next, the flow-balanced radial-based ML model for network traffic categorization uses the relevant extracted network traffic features. With the aid of flow, the balanced radial basis function reduces the influence of noise due to distinct network flow. This helps to improve accuracy and minimize latency. Due to this, better precision and recall is ensured. Performance of our method has been evaluated utilizing a scheme using an SDN traffic dataset. The results show that our method classifies the network traffic with high classification accuracy and minimum latency, ensuring better precision and recall.

Dynamic evolution of frontal-temporal network connectivity in temporal lobe epilepsy: A magnetoencephalography study.

Temporal lobe epilepsy (TLE) frequently involves an intricate, extensive epileptic frontal-temporal network. This study aimed to investigate the interactions between temporal and frontal regions and the dynamic patterns of the frontal-temporal network in TLE patients with different disease durations. The magnetoencephalography data of 36 postoperative seizure-free patients with long-term follow-up of at least 1 year, and 21 age- and sex-matched healthy subjects were included in this study. Patients were initially divided into LONG-TERM (n = 18, DURATION >10 years) and SHORT-TERM (n = 18, DURATION ≤10 years) groups based on 10-year disease duration. For reliability, supplementary analyses were conducted with alternative cutoffs, creating three groups: 0 < DURATION ≤7 years (n = 11), 7 < DURATION ≤14 years (n = 11), and DURATION >14 years (n = 14). This study examined the intraregional phase-amplitude coupling (PAC) between theta phase and alpha amplitude across the whole brain. The interregional directed phase transfer entropy (dPTE) between frontal and temporal regions in the alpha and theta bands, and the interregional cross-frequency directionality (CFD) between temporal and frontal regions from the theta phase to the alpha amplitude were further computed and compared among groups. Partial correlation analysis was conducted to investigate correlations between intraregional PAC, interregional dPTE connectivity, interregional CFD, and disease duration. Whole-brain intraregional PAC analyses revealed enhanced theta phase-alpha amplitude coupling within the ipsilateral temporal and frontal regions in TLE patients, and the ipsilateral temporal PAC was positively correlated with disease duration (r = 0.38, p <.05). Interregional dPTE analyses demonstrated a gradual increase in frontal-to-temporal connectivity within the alpha band, while the direction of theta-band connectivity reversed from frontal-to-temporal to temporal-to-frontal as the disease duration increased. Interregional CFD analyses revealed that the inhibitory effect of frontal regions on temporal regions gradually increased with prolonged disease duration (r = -0.36, p <.05). This study clarified the intrinsic reciprocal connectivity between temporal and frontal regions with TLE duration. We propose a dynamically reorganized triple-stage network that transitions from balanced networks to constrained networks and further develops into imbalanced networks as the disease duration increases.

BFFN: A novel balanced feature fusion network for fair facial expression recognition

Facial expression recognition (FER) technology has become increasingly mature and applicable in recent years. However, it still suffers from the bias of expression class, which can lead to unfair decisions for certain expression classes in applications. This study aims to mitigate expression class bias through both pre-processing and in-processing approaches. First, we analyze the output of existing models and demonstrate the existence of obvious class bias, particularly for underrepresented expressions. Second, we develop a class-balanced dataset constructed through data generation, mitigating unfairness at the data level. Then, we propose the Balanced Feature Fusion Network (BFFN), a class fairness-enhancing network. The BFFN mitigates the class bias by adding facial action units (AU) to enrich expression-related features and allocating weights in the AU feature fusion process to improve the extraction ability of underrepresented expression features. Finally, extensive experiments on datasets (RAF-DB and AffectNet) provide evidence that our BFFN outperforms existing FER models, improving the fairness by at least 16%.

Feature balanced re-enhanced network with multi-factor margin loss for long-tailed visual recognition

Real-world data often exhibits a long-tailed distribution, where the number of training samples for head classes far exceeds that of tail classes. This class imbalance phenomenon poses significant challenges for training deep neural networks. Existing class-aware loss methods typically focus only on the numerical relationship between class samples, blindly favoring the optimization of tail classes during the process, while neglecting the difficulty of samples and the similarity between the current class and other classes. To this end, relying only on the number relationship can easily lead to over-fitting of tail classes, thereby failing to fully utilize the potential information in the data. Therefore, we propose the Multi-Factor Margin Loss (MFMLoss), which consists of positive margin loss and negative margin loss. MFMLoss comprehensively considers three factors at three levels: overall, class, and sample: (1) quantitative relationships, (2) inter-class similarity relationships, and (3) sample recognition difficulty. The combined consideration of these three factors enables the model to pay more attention to confusing classes and difficult samples during the training process, rather than solely on tail classes, thus achieving optimization from coarse-grained to fine-grained. To further mitigate the negative impact of the imbalance between head and tail classes on feature learning, we design a new network architecture, called F-BREN. F-BREN consists of two components: the feature balancing network and the feature re-enhancement network. The former is trained with negative margin loss, which reduces the recognizability of easy samples. The latter is trained with positive margin loss, using positive margin to give more attention to hard samples, thus balancing the model’s attention to all samples. We conducted extensive experiments on four long-tailed benchmark datasets: CIFAR10-LT, CIFAR100-LT, ImageNet-LT and iNaturalist 2018, comparing the recognition accuracy of our method with eight state-of-the-art methods. The experimental results demonstrate that our proposed method outperforms the eight compared methods.

K-filter-based distributed adaptive bipartite tracking for nonlinear multi-agent systems with unknown control direction

This paper investigates K-filter-based distributed adaptive output-feedback bipartite tracking problem for multi-agent systems with unknown control direction. The agents are partitioned into two groups, where agents belonging to the same group are cooperative while agents belonging to the distinct group are competitive. Firstly, the unmeasurable state variables of the system under unknown control direction are estimated through K-filters. Then, a neighbourhood error is introduced to transform a structurally balanced network into a cooperative network. Based on K-filters, neighbourhood error, Nussbaum-type functions and dynamic surface control, a distributed adaptive output-feedback control protocol is designed to achieve the bipartite tracking. It is shown that all signals in the resulting closed-loop systems are semi-global bounded and the output of each agent can track the reference signal provided that the network is structurally balanced and contains a directed spanning tree. Finally, a numerical example is given to verify the effectiveness of the proposed method.

The recovery of parabolic avalanches in spatially subsampled neuronal networks at criticality

Scaling relationships are key in characterizing complex systems at criticality. In the brain, they are evident in neuronal avalanches—scale-invariant cascades of neuronal activity quantified by power laws. Avalanches manifest at the cellular level as cascades of neuronal groups that fire action potentials simultaneously. Such spatiotemporal synchronization is vital to theories on brain function yet avalanche synchronization is often underestimated when only a fraction of neurons is observed. Here, we investigate biases from fractional sampling within a balanced network of excitatory and inhibitory neurons with all-to-all connectivity and critical branching process dynamics. We focus on how mean avalanche size scales with avalanche duration. For parabolic avalanches, this scaling is quadratic, quantified by the scaling exponent, χ = 2, reflecting rapid spatial expansion of simultaneous neuronal firing over short durations. However, in networks sampled fractionally, χ is significantly lower. We demonstrate that applying temporal coarse-graining and increasing a minimum threshold for coincident firing restores χ = 2, even when as few as 0.1% of neurons are sampled. This correction crucially depends on the network being critical and fails for near sub- and supercritical branching dynamics. Using cellular 2-photon imaging, our approach robustly identifies χ = 2 over a wide parameter regime in ongoing neuronal activity from frontal cortex of awake mice. In contrast, the common ‘crackling noise’ approach fails to determine χ under similar sampling conditions at criticality. Our findings overcome scaling bias from fractional sampling and demonstrate rapid, spatiotemporal synchronization of neuronal assemblies consistent with scale-invariant, parabolic avalanches at criticality.

A dynamic balanced physics-informed neural network for solving partial differential equations

A dynamic balanced physics-informed neural network for solving partial differential equations

A Efficient and Accurate UAV Detection Method Based on YOLOv5s

Due to the limited computational resources of portable devices, target detection models for drone detection face challenges in real-time deployment. To enhance the detection efficiency of low, slow, and small unmanned aerial vehicles (UAVs), this study introduces an efficient drone detection model based on YOLOv5s (EDU-YOLO), incorporating lightweight feature extraction and balanced feature fusion modules. The model employs the ShuffleNetV2 network and coordinate attention mechanisms to construct a lightweight backbone network, significantly reducing the number of model parameters. It also utilizes a bidirectional feature pyramid network and ghost convolutions to build a balanced neck network, enriching the model’s representational capacity. Additionally, a new loss function, EloU, replaces CIoU to improve the model’s positioning accuracy and accelerate network convergence. Experimental results indicate that, compared to the YOLOv5s algorithm, our model only experiences a minimal decrease in mAP by 1.1%, while reducing GFLOPs from 16.0 to 2.2 and increasing FPS from 153 to 188. This provides a substantial foundation for networked optoelectronic detection of UAVs and similar slow-moving aerial targets, expanding the defensive perimeter and enabling earlier warnings.

The rigid-flexible balanced molecular crosslinking network transition interface: An effective strategy for improving the performance of bamboo fibers/poly(butadiene succinate-co-butadiene adipate) biocomposites

The poor interfacial compatibility of natural fiber-reinforced polymer composites has become a major challenge in the development of industry-standard high-performance composites. To solve this problem, this study constructs a novel rigid-flexible balanced molecular crosslinked network transition interface in composites. The interface improves the interfacial compatibility of the composites by balancing the stiffness and strength of the fibers and the matrix， effectively improving the properties of the composites. The flexural strength and flexural modulus of the composites were enhanced by 38 % and 44 %, respectively. Water absorption decreased by 30 %. The initial and maximum thermal degradation temperatures increased by 20 °C and 16 °C, respectively. The maximum storage modulus increased by 316 %. Furthermore, the impact toughness was elevated by 41 %, attributed to the crosslinked network's efficacy in absorbing and dissipating externally applied energy. This innovative approach introduces a new theory of interfacial reinforcement compatibility, advancing the development of high-performance and sustainable biocomposites.

Infrastructure-assisted 3D detection networks based on camera-lidar early fusion strategy

Target detection in autonomous driving is a pivotal domain of the current research focus. The process fundamentally relies on on-board sensors tasked to identify proximate objects. However, long-range perception instability and the partial obstruction by traffic participants further complicate the challenge. These factors collectively affected the effectiveness of targeted detection from the cooperative vehicle-infrastructure system (CVIS) and urgently need to be addressed. Starting from infrastructure-side assisted detection, we propose a 3D detection network based on multi-sensor sensing. Our approach consists of three sub-networks: Diversity Balanced Feature Fusion Network (MRBNeXt), Early Multimodal Fusion Network (VBRFusion), and TwoStage Lightweight Detection Network (TSL). MRBNeXt focuses on extracting raw images fused into multilevel semantic representations to address the drawbacks of needing a rich feature-level representation; VBRFusion proposes a two-branch structure that acts on the point cloud voxelization to aggregate high-dimensional features. The point features are mapped to the sampled graphical semantic features via the coordinate features to complete the early fusion and thus improve the feature quality of a single modality. In the proposed area network, TSL enhances the sensory field processing of multidimensional features using the context-aware module in a two-stage approach to achieve fast target recognition at different scale levels. Finally, we perform comparative ablation experiments on the DAIR-V2X vehicle-infrastructure dataset. The results validated our approach and demonstrated its effectiveness and enhancement in detection accuracy at the infrastructure end compared to current state-of-the-art methods. This improvement significantly boosted the performance of 3D target detection tasks in complex traffic scenarios and provided a more robust justification for the subsequent development of vehicle-side-infrastructure-side collaborative 3D target detection.

Two-stage metaheuristic for reliable and balanced network function virtualization-enabled networks

Two-stage metaheuristic for reliable and balanced network function virtualization-enabled networks

Balanced bandpass filter with common-mode reflectionless and linear-phase characteristics based on negative group delay circuit

In this paper, a balanced bandpass filter (BPF) with common-mode reflectionless and linear-phase characteristics is proposed, which consists of a balanced 3-order bandpass filtering network and two common-mode absorptive networks to obtain the common-mode reflectionless and suppression performance. The negative group delay circuits are loaded on the proposed balanced BPF to realize the full-passband linear-phase filtering performance. The flatter the group delay of the BPF (the fluctuation of the group delay within the passband tends to 0), the better the phase-frequency linearity, which will realize the undistortion transmission of the communication systems. For demonstration, a balanced linear-phase BPF prototype operating at 1.5 GHz is designed, simulated, and manufactured. The results demonstrate that the group delay fluctuation is reduced from 1.20 ns to 0.11 ns by 90.8 %, the differential-mode insertion loss is 1.04 dB, the minimum common-mode suppression is 59.3 dB within the passband and the common-mode reflectionless band is 1.22–1.91 GHz. The developed topology requires high precision in the processing, and the proposed balanced BPF can be applied to digital microwave communication systems to improve the transmission quality and reduce the bit error rate.

Intrinsic dynamics of randomly clustered networks generate place fields and preplay of novel environments.

During both sleep and awake immobility, hippocampal place cells reactivate time-compressed versions of sequences representing recently experienced trajectories in a phenomenon known as replay. Intriguingly, spontaneous sequences can also correspond to forthcoming trajectories in novel environments experienced later, in a phenomenon known as preplay. Here, we present a model showing that sequences of spikes correlated with the place fields underlying spatial trajectories in both previously experienced and future novel environments can arise spontaneously in neural circuits with random, clustered connectivity rather than pre-configured spatial maps. Moreover, the realistic place fields themselves arise in the circuit from minimal, landmark-based inputs. We find that preplay quality depends on the network's balance of cluster isolation and overlap, with optimal preplay occurring in small-world regimes of high clustering yet short path lengths. We validate the results of our model by applying the same place field and preplay analyses to previously published rat hippocampal place cell data. Our results show that clustered recurrent connectivity can generate spontaneous preplay and immediate replay of novel environments. These findings support a framework whereby novel sensory experiences become associated with preexisting "pluripotent" internal neural activity patterns.

Development of the Fetal Brain Corticocortical Structural Network during the Second-to-Third Trimester Based on Diffusion MRI.

During the second-to-third trimester, the neuronal pathways of the fetal brain experience rapid development, resulting in the complex architecture of the interwired network at birth. While diffusion MRI-based tractography has been employed to study the prenatal development of structural connectivity network (SCN) in preterm neonatal and postmortem fetal brains, the in utero development of SCN in the normal fetal brain remains largely unknown. In this study, we utilized in utero dMRI data from human fetuses of both sexes between 26 and 38 gestational weeks to investigate the developmental trajectories of the fetal brain SCN, focusing on intrahemispheric connections. Our analysis revealed significant increases in global efficiency, mean local efficiency, and clustering coefficient, along with significant decrease in shortest path length, while small-worldness persisted during the studied period, revealing balanced network integration and segregation. Widespread short-ranged connectivity strengthened significantly. The nodal strength developed in a posterior-to-anterior and medial-to-lateral order, reflecting a spatiotemporal gradient in cortical network connectivity development. Moreover, we observed distinct lateralization patterns in the fetal brain SCN. Globally, there was a leftward lateralization in network efficiency, clustering coefficient, and small-worldness. The regional lateralization patterns in most language, motor, and visual-related areas were consistent with prior knowledge, except for Wernicke's area, indicating lateralized brain wiring is an innate property of the human brain starting from the fetal period. Our findings provided a comprehensive view of the development of the fetal brain SCN and its lateralization, as a normative template that may be used to characterize atypical development.

Event-Triggered Adaptive Containment Control for Heterogeneous Stochastic Nonlinear Multiagent Systems.

This article investigates the event-triggered adaptive containment control problem for a class of stochastic nonlinear multiagent systems with unmeasurable states. A stochastic system with unknown heterogeneous dynamics is established to describe the agents in a random vibration environment. Besides, the uncertain nonlinear dynamics are approximated by radial basis function neural networks (NNs), and the unmeasured states are estimated by constructing the NN-based observer. In addition, the switching-threshold-based event-triggered control method is adopted with the hope of reducing communication consumption and balancing system performance and network constraints. Moreover, we develop the novel distributed containment controller by utilizing the adaptive backstepping control strategy and the dynamic surface control (DSC) approach such that the output of each follower converges to the convex hull spanned by multiple leaders, and all signals of the closed-loop system are cooperatively semi-globally uniformly ultimately bounded in mean square. Finally, we verify the efficiency of the proposed controller by the simulation examples.

Surface defect identification method for hot-rolled steel plates based on random data balancing and lightweight convolutional neural network

Surface defect identification method for hot-rolled steel plates based on random data balancing and lightweight convolutional neural network

EB-Net: an efficient balanced network for accuracy and speed of remote sensing detection

ABSTRACT Remote sensing detection is a difficult task that requires not only identifying objects with complex background, quality, and angle issues, but also being lightweight enough to be carried by edge devices. In real remote sensing scenarios, achieving accurate, fast, and low-resource-consumption automated detection remains a significant challenge. Therefore, this paper proposes an efficient balanced network (EB-Net) for real remote sensing devices. First, a dynamic sparse attention (DSA) mechanism is proposed and has been proven with high performance via the complexity analysis. In addition, a new dynamic sparse transformer (DSFormer) is constructed using DSA, which enhances feature information and adapts to image resolution by self-attention and multi-headed attention, and achieves more flexible computation by random sampling. Then, three versions of discrete distribution IoU (DDIoU) are defined for adapting various scenarios and tasks in remote sensing, and this loss function makes the model achieve high accuracy and more lightweight. Finally, to make the model more lightweight, a cropped SPPF (CroSPPF) is presented, which significantly improves the computational efficiency by the lightweight of the sequence structure and activation function. Ablation experiments are conducted on the NWPU VHR-10 dataset, and demonstrate the effectiveness of the proposed methods. Numerous comparisons with state-of-the-art detectors are conducted on the NWPU VHR-10, RSOD and DOTA datasets. The experiments show that EB-Net outperforms state-of-the-art remote sensing detection models in comprehensive performance and achieves end-to-end accurate and lightweight detection.