Random Loss Research Articles

Affine Formation Maneuver Control for NUSVs: An Anti-Competing Interaction Solution With Random Packet Losses

Affine Formation Maneuver Control for NUSVs: An Anti-Competing Interaction Solution With Random Packet Losses

A Variational Bayesian Adaptive Kalman Filter for the Random Losses Problem of Sensor Packet

A Variational Bayesian Adaptive Kalman Filter for the Random Losses Problem of Sensor Packet

Unusual rearrangements of mitogenomes in Diptera revealed by comparative analysis of 135 tachinid species (Insecta, Diptera, Tachinidae)

The Tachinidae is one of the most speciose families in Diptera, and the exclusively parasitoid species play an important role in regulating populations of many herbivorous insects in ecosystems, including many agricultural pests. To better comprehend the characteristics and evolution of the mitochondrial genome for the Tachinidae, we are adding a massive amount of new molecular data by assembling the mitogenomes for 71 genera and 135 species from all four tachinid subfamilies through next-generation sequencing, and we are presenting the most comprehensive mitogenomic phylogenetic analysis of this family so far. Extensive rearrangements observed in the mitogenome of Admontia podomyia (Exoristinae) are unique for the entire suborder Cyclorrhapha. The rearrangement pattern suggests that the process involved a tandem duplication of the complete mitogenome, followed by both random and nonrandom loss of one copy of each gene. Additionally, five minor mitogenome rearrangements are discovered and described in three subfamilies. We present the largest species-level phylogenetic hypothesis for Tachinidae to date, based on mitogenomes of 152 species of Tachinidae, representing all four subfamilies and with five non-tachinid outgroups. Our analyses support the monophyly of the Tachinidae and most tribes and genera were recovered with good support, but the higher-level phylogenetic relationships within Tachinidae were poorly resolved, indicating that mitogenome data alone are not enough to unambiguously resolve the deeper phylogenetic relationships within Tachinidae.

The role of ion migration, octahedral tilt, and the A-site cation on the instability of Cs1-xFAxPbI3

Organic-inorganic hybrid perovskites are promising materials for the next generation photovoltaics and optoelectronics; however, their practical application has been hindered by poor structural stability mainly caused by ion migration and external stimuli. Understanding the mechanism(s) of ion migration and structure decomposition is thus critical. Here we observe the sequence of structural changes at the atomic level that precede structural decomposition in the technologically important Cs1-xFAxPbI3 using ultralow dose transmission electron microscopy. We find that these changes differ, depending upon the A-site composition. Initially, there is a random loss of FA+, complemented by the loss of I-. The remaining FA+ and I- ions then migrate, unit cell by unit cell, into an ordered and more stable phase with a √2 x √2 superstructure. Further ion loss is accompanied by A-site dependent octahedral tilt modes and associated tetragonal phases with different stabilities. These observations of the loss of FA+/I- ion pairs, ion migration, octahedral tilt modes, and the role of the A-cation, provide insights into the atomic-scale structural mechanisms that drive and block ion loss and ion migration, opening pathways to inhibit ion loss, migration and improve structural stability.

Random epipolar constraint loss functions for supervised optical flow estimation

The majority of supervised models estimate optical flow through minimizing the numerical difference between the predicted flow and the ground truth, resulting in the loss of positional and geometric characteristics of the calculated flow fields. In addition, these models require a large number of parameters and high computational cost when computing optical flow. To address these issues, this paper presents a novel loss function and a lightweight framework for optical flow estimation. The proposed loss function, called the random epipolar constraint loss function (RECLoss), incorporates epipolar geometry into supervised optimization to transform the numerical difference into geometry constraint. The RECLoss can make the optical flow estimation models more effective and enhance their generalization abilities. Moreover, the design of RECLoss is more interpretable and the estimated optical flow fields from RECLoss have clearly defined mathematical meanings. A lightweight recurrent neural network for optical flow estimation (LRFlow) that balances computational cost and estimation accuracy is also proposed. The LRFlow, containing only 3.0M parameters, consists of a feature extractor, a correlation matching module, and an iterative update unit. The proposed lightweight network achieves state-of-the-art results compared to all other lightweight networks on the challenging MPI-Sintel and KITTI2015 datasets. The effectiveness of RECLoss in improving the accuracy of LRFlow and other state-of-the-art methods such as RAFT and GMA has been validated through extensive experiments. The source code of the project are available at https://github.com/Eryo-iPython/RECLoss.

BrainGridNet: A two-branch depthwise CNN for decoding EEG-based multi-class motor imagery

Brain–computer interfaces (BCIs) based on motor imagery (MI) enable the disabled to interact with the world through brain signals. To meet demands of real-time, stable, and diverse interactions, it is crucial to develop lightweight networks that can accurately and reliably decode multi-class MI tasks. In this paper, we introduce BrainGridNet, a convolutional neural network (CNN) framework that integrates two intersecting depthwise CNN branches with 3D electroencephalography (EEG) data to decode a five-class MI task. The BrainGridNet attains competitive results in both the time and frequency domains, with superior performance in the frequency domain. As a result, an accuracy of 80.26 percent and a kappa value of 0.753 are achieved by BrainGridNet, surpassing the state-of-the-art (SOTA) model. Additionally, BrainGridNet shows optimal computational efficiency, excels in decoding the most challenging subject, and maintains robust accuracy despite the random loss of 16 electrode signals. Finally, the visualizations demonstrate that BrainGridNet learns discriminative features and identifies critical brain regions and frequency bands corresponding to each MI class. The convergence of BrainGridNet’s strong feature extraction capability, high decoding accuracy, steady decoding efficacy, and low computational costs renders it an appealing choice for facilitating the development of BCIs.

Hierarchical multimodal-fusion of physiological signals for emotion recognition with scenario adaption and contrastive alignment

The lack of complementary affective responses from both the central and peripheral nervous systems could limit the performance of emotion recognition with the single-modal physiological signal. However, when integrating multimodalities, a direct fusion may ignore the heterogeneous nature of multiple feature domains from one modality to another. Besides, there is a risk that the distribution of the multimodal physiological responses may vary across different affective scenarios for stimulating an identical emotional category. The inter-individual variation may also increase due to the superposition of the biometric information from the multimodal features. To tackle these issues, we present a hierarchical multimodal network for robust heterogeneous physiological representations (RHPRNet). First, we applied a spatial-frequency pattern extractor to identify the electroencephalogram (EEG) representations in both the spatial and frequency domains. Next, inter-domain and inter-modality affective encoders are separately applied to the statistic-complexity EEG features and multimodal peripheral features, respectively. All the learned representations are integrated via a hierarchical fusion module. To model the multi-peak patterns stimulated by different affective scenarios, we designed a scenario-adapting pretraining stage. A random contrastive training loss was also applied to mitigate the inter-individual variance. In the end, we performed adequate experiments to examine the performance of the RHPRNet based on three publicly available multimodal databases combined with two validation approaches.

Observer‐based non‐fragile event‐triggered control of extra‐corporeal blood circulation process in finite‐time interval

AbstractAn observer‐based event‐triggered non‐fragile control problem for regulating the levels of oxygen and carbon dioxide (O and CO) in patients blood during the extra‐corporeal blood circulation (ECC) process with random communication packet losses is explored in this work. The main intention of this work is to design an observer‐based event‐triggered and non‐fragile controller for ensuring the finite‐time boundedness criterion for the resulting system with mixed and passivity performance index over a finite interval of time. In addition to the limited bandwidth of the channels, such random packet losses might occur simultaneously between the sensors and the controllers, and between the controllers and the actuators. Unlike most studies in the literature, we have found that, since the controller‐actuator link has packet dropouts, the observer's control input differs from that of the plant. The sufficient conditions derived are presented in a linear matrix inequalities frame work. Finally, simulation results are used to validate the efficacy of the proposed control method. The results show that the proposed observer‐based event‐triggered non‐fragile controller keeps the blood gases within physiological limits during a stable ECC process.

Random forest-based multi-hazard loss estimation using hypothetical data at seismic and tsunami monitoring networks

This article presents a novel approach to estimate multi-hazard loss in a post-event situation, resulting from cascading earthquake and tsunami events with machine learning for the first time. The proposed methodology combines the power of random forest (RF) with data that are simulated at seismic and tsunami monitoring locations. The RF model is well-suited for predicting highly nonlinear multi-hazard loss because of its nonparametric regression and ensemble learning capabilities. The study targets the cities of Iwanuma and Onagawa in Tohoku, Japan, where seismic and tsunami monitoring networks have been deployed. To encompass a diverse range of future multi-hazard loss estimation, an RF model is constructed based on 4000 simulated earthquake events with peak ground velocity and tsunami wave amplitude captured at ground-motion monitoring sites and offshore wave monitoring sensors, respectively. The incorporation of 10 ground-motion monitoring sites and five offshore wave monitoring sensors significantly enhances the model’s forecasting power, leading to a notable 60% decrease in mean squared error and 20% increase in the R 2 value compared to scenarios where no monitoring sensors are utilized. By harnessing the capabilities of RF and leveraging detailed sensing data, RF achieves R 2 values over 90%, which can contribute to enhanced disaster risk management.

Mixed Event-Triggered Output Regulation for Networked Switched Systems With Unstable Switching Dynamics Under Long-Duration DoS Attacks.

In this article, the event-triggered output regulation problem (EORP) under the denial-of-service (DoS) attacks is considered for networked switched systems (NSSs) with unstable switching dynamics (USDs). The USDs here refer to the unsolvable output regulation of each subsystem and the destabilization at partial switching instants, which indicates that the Lyapunov function does not decrease monotonically in activation intervals of each subsystem and increases at partial switching instants. First, long-duration DoS attacks (LDDAs) are considered, where LDDAs imply that their duration may be longer than the total dwell time (DT) of several adjacent activated subsystems. By imposing constraints at switching instants, consecutive asynchronous subsystem switching caused by LDDAs and USDs is allowed, that is, the subsystem switches several times but the controller switching is blocked by LDDAs and controllers fail to switch correspondingly. Second, mixed event-triggered mechanisms (ETMs), combining event-triggered conditions and periodic sampling conditions, are designed to reduce network burden under LDDAs and improve system performance subject to destabilizing switching. Then, an improved DT for switching signal permits irregular arrangement of destabilizing and stabilizing switching and is more suitable for NSSs subject to LDDAs. Moreover, sufficient conditions ensure the solvability of EORP for NSSs with USDs under LDDAs, network-induced delays, random packet losses, and packet disorders. Finally, a switched RLC circuit shows the feasibility of the proposed method.

The declines of heterogeneity and stability in diatom communities are associated with human activity.

Anthropogenic forcing caused the biodiversity loss and stability decline of communities. There is still controversy over whether the decline in biodiversity will lead to a decrease in community stability. The stability of biological communities is related to both biodiversity and structure, and this paper aims to reveal the human impacts on diatom communities' biodiversity and structure. We studied the richness, β-diversity and network distance of diatom communities in Qinghai-Xizang, Yunnan-Sichuan and Lower Yangtze River Basin, China through empirical dataset and simulation method. The results showed that the diatoms richness in the Qinghai-Xizang and the Yunnan-Sichuan region was lower and the network distance was higher than that of the Lower Yangtze River Basin. β-diversity in the Lower Yangtze River Basin was the lowest and the diatom network distance responds negatively to human population densities in China. The simulation showed that the network distance kept constant during random species loss, and declined while specialist species were lost or replaced by generalist species. The results suggested diatom communities' homogeneity and stability decline were associated with human activities. Human impacts may cause biodiversity loss targeted to specialist species or no biodiversity loss while generalist species replace those specialist species. This study showed that how diversity changes determined ecological stability depends on the type of species changes.

Ultra‐Low‐Crosstalk Silicon Arrayed‐Waveguide Grating (De)multiplexer with 1.6‐nm Channel Spacing

AbstractA silicon arrayed‐waveguide grating (AWG) with 1.6‐nm channel spacing is proposed and realized with high performances for dense wavelength‐division (de)multiplexing systems. For the present AWG, the arrayed waveguides are broadened uniformly to be far beyond the singlemode regime, so that random phase errors and propagation loss are minimized even without any additional requirement for the fabrication process. Particularly, Euler‐bends are introduced for the arrayed waveguides to shrink the device footprint and suppress any excitation of higher‐order modes. As an example, a 16 × 16 AWG (de)multiplexer is realized with a compact footprint of 600 × 800 µm2 by using single‐etched 220‐nm‐thick silicon photonic waveguides. When using the fabricated AWG to demultiplex the channels launched from the central input port, the excess loss and the adjacent‐channel crosstalk of the central output port are ≈2.2 and ≈−31.7 dB, respectively. Such an AWG is one of the best among silicon implementations.

Optimal Insurance: Dual Utility, Random Losses, and Adverse Selection

We study a generalization of the classical monopoly insurance problem under adverse selection (see Stiglitz 1977) where we allow for a random distribution of losses, possibly correlated with the agent’s risk parameter that is private information. Our model explains patterns of observed customer behavior and predicts insurance contracts most often observed in practice: these consist of menus of several deductible-premium pairs or menus of insurance with coverage limits–premium pairs. A main departure from the classical insurance literature is obtained here by endowing the agents with risk-averse preferences that can be represented by a dual utility functional (Yaari 1987). (JEL D81, D82, D86, D91, G22)

Adaptive fault-tolerant control on complex networks with semi-Markov jump topology under random packet loss

ABSTRACT In this paper, semi-Markov process and adaptive fault-tolerant law are used to study the synchronization of complex networks in the presence of topology mutations and actuator failures. Firstly, the semi-Markov jump process, which is more practical than the traditional Markov jump process, is used to simulate the topology change of complex networks. Secondly, by designing adaptive fault-tolerant controller, the influence of actuator fault and interference on the system is eliminated. Then, considering the random packet loss caused by network channel congestion and limited bandwidth resources, a control strategy is developed to achieve convergence of the synchronisation error system. Finally, a numerical example and a practical example of Chua's circuit are given to verify the validity of the proposed theory.

3-D Localization of Multiagent Systems Under Random Environments Based on Iterative Learning

This work addresses the localization of dynamic multi-agent systems (MASs) in three-dimensional (3-D) space under random environments (i.e., random packet loss, measurement noise and simultaneous communication noise). It is more practical yet challenging to deal with than static systems under ideal environments. Barycentric coordinates based on the relative-distance are used to characterize the positions between agents. The sign coefficients are introduced to guarantee that the sum of barycentric coordinates is equal to one, thus eliminating the constraint that the agent needs to be located in the convex hull. Using the newly designed iteration-varying gains, a robust distributed estimation method based on iterative learning is proposed. The key to accurate localization is to construct an appropriate approximator via introducing two iterative-varying gains into the localization scheme. The convergence in the sense of mathematical expectation is proved. The simulation examples and experimental results verify the effectiveness of the proposed approach.

Fossilization can mislead analyses of phenotypic disparity.

Analyses of morphological disparity can incorporate living and fossil taxa to facilitate the exploration of how phenotypic variation changes through time. However, taphonomic processes introduce non-random patterns of data loss in fossil data and their impact on perceptions of disparity is unclear. To address this, we characterize how measures of disparity change when simulated and empirical data are degraded through random and structured data loss. We demonstrate that both types of data loss can distort the disparity of clades, and that the magnitude and direction of these changes varies between the most commonly employed distance metrics and disparity indices. The inclusion of extant taxa and exceptionally preserved fossils mitigates these distortions and clarifies the full extent of the data lost, most of which would otherwise go uncharacterized. This facilitates the use of ancestral state estimation and evolutionary simulations to further control for the effects of data loss. Where the addition of such reference taxa is not possible, we urge caution in the extrapolation of general patterns in disparity from datasets that characterize subsets of phenotype, which may represent no more than the traits that they sample.

Assessing the Sensitivity of OCT-A Retinal Vasculature Metrics.

The purpose of this study was to examine the sensitivity of quantitative metrics of the retinal vasculature derived from optical coherence tomography angiography (OCT-A) images. Full retinal vascular slab OCT-A images were obtained from 94 healthy participants. Capillary loss, at 1% increments up to 50%, was simulated by randomly removing capillary segments (1000 iterations of randomized loss for each participant at each percent loss). Thirteen quantitative metrics were calculated for each image: foveal avascular zone (FAZ) area, vessel density, vessel complexity index (VCI), vessel perimeter index (VPI), fractal dimension (FD), and parafoveal intercapillary area (PICA) measurements with and without the FAZ (mean PICA, summed PICA, PICA regularity, and PICA standard deviation [PICA SD]). The sensitivity of each metric was calculated as the percent loss at which 80% of the iterations for a participant fell outside of two standard deviations from the sample's normative mean. The most used OCT-A metrics, FAZ area and vessel density, were not significantly different from normative values until 27.69% and 16.00% capillary loss, respectively. Across the remaining metrics, metric sensitivity ranged from 6.37% (PICA SD without FAZ) to 39.78% (Summed PICA without FAZ). The sensitivity of vasculature metrics for detecting random capillary loss varies substantially. Further efforts simulating different patterns of capillary loss are needed for comparison. Additionally, mapping the repeatability of metrics over time in a normal population is needed to further define metric sensitivity. Quantitative metrics vary in their ability to detect vascular abnormalities in OCT-A images. Metric choice in screening studies will need to balance expected capillary abnormalities and the quality of the OCT-A images being used.

Random Cycle Loss and Its Application to Voice Conversion

Speech disentanglement aims to decompose independent causal factors of speech signals into separate codes. Perfect disentanglement benefits to a broad range of speech processing tasks. This paper presents a simple but effective disentanglement approach based on cycle consistency loss and random factor substitution. This leads to a novel random cycle (RC) loss that enforces analysis-and-resynthesis consistency, a main principle of reductionism. We theoretically demonstrate that the proposed RC loss can achieve independent codes if well optimized, which in turn leads to superior disentanglement when combined with information bottleneck (IB). Extensive simulation experiments were conducted to understand the properties of the RC loss, and experimental results on voice conversion further demonstrate the practical merit of the proposal. Source code and audio samples can be found on the webpage http://rc.cslt.org.

Resistance of aerobic granular sludge microbiomes to periodic loss of biomass

Granular sludge is a biofilm process used for wastewater treatment which is currently being implemented worldwide. It is important to understand how disturbances affect the microbial community and performance of reactors. Here, two acetate-fed replicate reactors were inoculated with acclimatized sludge and the reactor performance, and the granular sludge microbial community succession were studied for 149 days. During this time, the microbial community was challenged by periodically removing half of the reactor biomass, subsequently increasing the food-to-microorganism (F/M) ratio. Diversity analysis together with null models show that overall, the microbial communities were resistant to the disturbances, observing some minor effects on polyphosphate-accumulating and denitrifying microbial communities and their associated reactor functions. Community turnover was driven by drift and random granule loss, and stochasticity was the governing ecological process for community assembly. These results evidence the aerobic granular sludge process as a robust system for wastewater treatment.

Asynchronous event-triggered [formula omitted][formula omitted] control for 2D Markov jump systems subject to networked random packet losses

This paper investigates the issue of asynchronous event-triggered control for two-dimensional Markov jump systems subject to networked random packet losses. The asynchronous behavior between the plant and controller is characterized by a hidden Markov model, and the packet loss behavior is described by a Bernoulli stochastic process. To relieve the communication pressure, an event-triggered strategy with independent triggering behavior in the horizontal and vertical directions is given. A sufficient condition is established to ensure stochastic stability and H∞ performance for the closed-loop system. Then, a method for designing an asynchronous event-triggered controller is developed through eliminating nonlinearities formed by the inverse of Lyapunov matrices. Moreover, based on convex optimization of the H∞ performance, an algorithm is proposed to find the maximum feasible domain of the event-triggered thresholds. Finally, a Darboux equation with Markov jump properties is applied to verify the effectiveness of the presented design method.