Observation Noise Research Articles

MKAN-MMI: empowering traditional medicine-microbe interaction prediction with masked graph autoencoders and KANs

The growing microbial resistance to traditional medicines necessitates in-depth analysis of medicine-microbe interactions (MMIs) to develop new therapeutic strategies. Widely used artificial intelligence models are limited by sparse observational data and prevalent noise, leading to over-reliance on specific data for feature extraction and reduced generalization ability. To address these limitations, we integrate Kolmogorov-Arnold Networks (KANs), independent subspaces, and collaborative decoding techniques into the masked graph autoencoder (Mask GAE) framework, creating an innovative MMI prediction model with enhanced accuracy, generalization, and interpretability. First, we apply Bernoulli distribution to randomly mask parts of the medicine-microbe graph, advancing self-supervised training and reducing noise impact. Additionally, the independent subspace technique enables graph neural networks (GNNs) to learn weights independently across different feature subspaces, enhancing feature expression. Fusing the multi-layer outputs of GNNs effectively reduces information loss caused by masking. Moreover, using KANs for advanced nonlinear mapping enhances the learnability and interpretability of weights, deepening the understanding of complex MMIs. These measures significantly enhanced the accuracy, generalization, and interpretability of our model in MMI prediction tasks. We validated our model on three public datasets with results showing that our model outperformed existing leading models. The relevant data and code are publicly accessible at: https://github.com/zhuoninnin1992/MKAN-MMI.

Computational processes of simultaneous learning of stochasticity and volatility in humans

Making adaptive decisions requires predicting outcomes, and this in turn requires adapting to uncertain environments. This study explores computational challenges in distinguishing two types of noise influencing predictions: volatility and stochasticity. Volatility refers to diffusion noise in latent causes, requiring a higher learning rate, while stochasticity introduces moment-to-moment observation noise and reduces learning rate. Dissociating these effects is challenging as both increase the variance of observations. Previous research examined these factors mostly separately, but it remains unclear whether and how humans dissociate them when they are played off against one another. In two large-scale experiments, through a behavioral prediction task and computational modeling, we report evidence of humans dissociating volatility and stochasticity solely based on their observations. We observed contrasting effects of volatility and stochasticity on learning rates, consistent with statistical principles. These results are consistent with a computational model that estimates volatility and stochasticity by balancing their dueling effects.

Bias and multiscale correction methods for variational state estimation

Data assimilation performance can be significantly impacted by biased noise in observations, altering the signal magnitude and introducing fast oscillations or discontinuities when the system lacks smoothness. To mitigate these issues, this paper employs variational state estimation using the so-called parametrized-background data-weak method. This approach relies on a background manifold parametrized by a set of constraints, enabling the state estimation by solving a minimization problem on a reduced-order background model, subject to constraints imposed by the input measurements. The proposed formulation incorporates a novel bias correction mechanism and a manifold decomposition that handles rapid oscillations by treating them as slow-decaying modes based on a two-scale splitting of the classical reconstruction algorithm. The method is validated in different examples, including the assimilation of biased synthetic data, discontinuous signals, and Doppler ultrasound data obtained from experimental measurements.

Speckle-free holography with a diffraction-aware global perceptual model

Computer-generated holography (CGH) based on neural networks has been actively investigated in recent years, and convolutional neural networks (CNNs) are frequently adopted. A convolutional kernel captures local dependencies between neighboring pixels. However, in CGH, each pixel on the hologram influences all the image pixels on the observation plane, thus requiring a network capable of learning long-distance dependencies. To tackle this problem, we propose a CGH model called Holomer. Its single-layer perceptual field is 43 times larger than that of a widely used 3×3 convolutional kernel, thanks to the embedding-based feature dimensionality reduction and multi-head sliding-window self-attention mechanisms. In addition, we propose a metric to measure the networks’ learning ability of the inverse diffraction process. In the simulation, our method demonstrated noteworthy performance on the DIV2K dataset at a resolution of 1920×1024, achieving a PSNR and an SSIM of 35.59 dB and 0.93, respectively. The optical experiments reveal that our results have excellent image details and no observable background speckle noise. This work paves the path of high-quality hologram generation.

Soil Moisture Detection and Linear Deceleration Control Strategy Enhancing Trenching Depth Precision and Stability for Rapeseed Sowing

Sowing depth significantly affects the germination of rapeseed, and different soil moisture conditions require corresponding sowing depths. However, most current trenching devices do not account for soil moisture content, and commonly used hydraulic or constant-force trenching equipment also exhibits deficiencies in stability and consistency. To address these challenges, this study developed an automatic depth adjustment control system based on soil moisture content. A soil moisture detection device and an innovative sliding mechanism that maintained the soil moisture sensor in a relatively stationary position relative to the soil during seeder movement were introduced. An automatic sowing depth adjustment device was designed to modulate the sowing depth. A control strategy that incorporated the Kalman filtering algorithm and linear deceleration equations was conducted. At an observation noise covariance matrix (Q/R) of 0.001, a deceleration range of 40 mm and a minimum speed of 10, the control system exhibited minimal overshoot (approximately 4%) and steady-state error (approximately 3.2 mm). It effectively adjusted the trenching depth while operating at speeds ranging from 2 to 3.6 km/h, successfully adapting to variations in soil topography. The system performance tests revealed that the control system adjustment time (ts) was 534 ms and the steady-state error remained within 1 mm. Under three different soil moisture content conditions, the sowing depth qualification rate and stability coefficients consistently surpassed 90% and 80%, respectively. This research offers a sowing depth adjustment control system based on soil moisture content, contributing to more precise depth regulation for rapeseed sowing.

Partially observable optimal control using exponential cost criterion

We consider a partially observable optimal control model in which the system and observation noises are correlated, and the cost criterion is in the form of an exponential of an integral. A new minimum principle criterion is derived and applied to compute the explicit solution of linear exponential Gaussian optimisation problem with correlated noises.

Three-dimensional cooperative guidance with impact angle constraints via value-policy decomposed multi-agent reinforcement learning

For a three-dimensional impact angle-constrained time-coordination guidance problem, a value-policy decomposed multi-agent twin delayed deep deterministic (VPD-MATD3) policy gradient algorithm is proposed in this paper, and a temporal-spatial cooperative guidance law is designed consequently. The derived cooperative engagement kinematics is formulated as a value-policy decomposed partially observable Markov decision process. Value decomposition and policy decomposition are correspondingly proposed to address the two bottlenecks of reward coupling and policy coupling in 3D guidance, thus improving the training process. The training environment with multiple uncertainties, such as observation noise, maneuverability perturbation, autopilot time lag, switching communication topology, and random initialization, is subsequently constructed. Time-energy type reward functions are designed, incorporating energy consumption as a direct optimization indicator via a penalty term. A long short-term memory network layer is inserted inside the policy networks to suppress the negative effect of observation noise. The policy training curves verify the significant effect of the proposed algorithm in improving training convergence. The policy testing results illustrate the robustness and generalization of the VPD-MATD3 cooperative guidance law and its ability to cope with randomly switched communication topologies. Moreover, the comparative testing further reveals its time-energy suboptimal property.

A Decoupled Feedback and Fast Norm Optimal Control approach for Normal and Superconducting RF Cavity disturbances compensation

Iterative learning controller (ILC)s of different types have been used in various engineering applications for system control in presence of repetitive disturbance, often encountered in batch control systems. A particle accelerator usually comprising of Radiofrequnecy (RF) cavities, when operated in pulsed mode processes particle beam similar to that in batch systems. However, an injected particle beam disturbs electric field inside RF cavity. In additions to beam loading disturbance encountered in all types of RF cavities, a superconducting cavity also encounters disturbance due to Lorentz force detuning. Under presence of such persisting disturbance/s, a Fast Norm Optimal ILC (FNOILC), helps achieve specified amplitude and phase stability of electric field, which results in beam of specified quality. A sophisticated control system helps achieve this objective. In this paper, FNOILC technique is explored for this purpose. Though, it has been reported for few accelerator facilities, in this work, a decoupling strategy is used in a feedback control loop, resulting into overall enhancement of control performance. Behavioral model of RF cavity is represented as dual input dual output system exhibiting two loops, found interacting due to nonzero cavity detuning. A decoupler is designed based on well known technique/s used in process control applications. Relative gain array element values are used to quantify interaction. In presence of beam loading disturbance, a FNOILC demonstrates an excellent control performance. A standard performance index has been used to assess control performance. It is found that use of decoupler in main feedback loop of RF cavity, enhances the resulting performance index. Besides, an ILC performance needs assessment with practical issues like model uncertainty, sensing delays, high detuning and observation noise. In this paper, most of conclusions are drawn based on large number of modeling and simulation studies carried out for lab prototype 650MHz cavity. Interesting studies like effects of low controller sampling frequency, feedback controller gain tuning, severe detuning, etc. are presented using a novel state plane representation. A zero phase filtering is simulated and proposed as an effective strategy when measurements have noise. This paper forms a case study which can be applied to any general system encountering repetitive disturbance. Additionally, paper also elaborates on FNOILC algorithm application for Lorentz force detuning compensation often encountered in superconducting cavity during pulse mode of operation.

Spot modelling through multi-band photometry. Analysis of V1298 Tau

Stellar activity is comprised of various phenomena, mainly spots and faculae. It is one of the main sources of noise in exoplanetary observations because it affects both spectroscopic and photometric observations. In studying young active planetary systems, we need to model the activity of the host stars to remove astrophysical noise from our observational data. We model the contribution of stellar spots in photometric observations. Through the use of multi-band photometry, we aim to extract the geometric properties of the spots and constrain their temperatures. We analysed multi-band photometric observations acquired with the 80 cm Marcon telescope of the Osservatorio Polifunzionale del Chianti of V1298 Tau, assuming the photometric modulation observed in different bands is attributed to cold spots. We constrained the effective temperature of the active regions present on the surface of V1298 Tau, resulting from a combination of spots and faculae. We tested our hypothesis on solar data, verifying that we successfully measured the size of the dominant active region and its averaged effective temperature.

Integrating low-cost GNSS and MEMS accelerometer for precise dynamic displacement monitoring

We assess the performance of the self-developed receiver coupling dual-frequency low-cost GNSS chipset and IMU MEMS sensor. The system also comprises software dedicated to vibration monitoring based on relative kinematic positioning in a post-processing mode. The observation assessment reveals high accuracy of low-cost GNSS and MEMS accelerometer data. The Septentrio Mosaic X5 chipset exhibits competitive to high-grade receivers’ phase observation noise, outperformance in code noise level and observation auto-correlation at a level that may be neglected even in high-rate applications. The evaluation of the MEMS accelerometer data proves that the sensor accuracy is adequate for vibration monitoring. With the shake table experiment, we examine the performance of vibration recovery. An advantage of the coupled solution over a GNSS-only one is documented with a 15 % reduction of the amplitude error. The backward smoothing of coupled solutions drives the next 20 % error drop. Finally, the high performance of the coupled solution based on low-cost GNSS & accelerometer is confirmed by a mean amplitude error of 0.4 mm. The frequency analyses indicate that both coupled and single-sensor solutions precisely detect the vibration frequencies; however, they also confirm the outperformance of the former over the latter. The power spectra of the recovered displacement time series accentuate the benefits of GNSS + accelerometer fusion and backward filtering for reducing displacement noise.

Box replication effects in weak lensing light-cone construction

Abstract Weak gravitational lensing simulations serve as indispensable tools for obtaining precise cosmological constraints. In particular, it is crucial to address the systematic uncertainties in theoretical predictions, given the rapid increase in galaxy numbers and the reduction in observational noise. Both on-the-fly and post-processing methods for constructing lensing light-cones encounter limitations due to the finite simulated volume, necessitating the replication of the simulation box to encompass the volume to high redshifts. To address this issue, our primary focus lies on investigating and quantifying the impact of box replication on the convergence power spectrum and higher-order moments of lensing fields. Subsequently, a univariate model is utilized to estimate the amplitude parameter A by fitting four statistics measured from partial-sky light-cones along specific angles, to the averaged result from random directions. The investigation demonstrates that the systematic bias stemming from the box replication phenomenon falls within the bounds of statistical errors for the majority of cases. However, caution should be exercised when considering high-order statistics on a small sky coverage (≲ 25 deg2). For this case, we have developed a code that facilitates the identification of optimal viewing angles for the light-cone construction. This code has been made publicly accessible https://github.com/czymh/losf.

Explainable, Deep Reinforcement Learning–Based Decision Making for Operations and Maintenance

This paper presents research that integrates condition monitoring and prognostics with decision making for nuclear power plant operations and maintenance aimed at reducing lifetime maintenance and repair costs. Additionally, a focal point of this research is to make the decisions explainable to operators, improving the trustworthiness of the decisions from what can be considered a black box model. In this work, we develop and evaluate an explainable, online asset management methodology to help reduce lifetime maintenance and repair costs. Using the latest advancements in condition monitoring, inventory management, deep reinforcement learning, and explainable artificial intelligence methods, we create a predictive maintenance methodology that can optimize the maintenance and spare part management of a repairable nuclear power plant system. To demonstrate these methods, preliminary studies were conducted on a representative maintenance system undergoing a stochastic degradation process that requires repairs or replacement to continue operation. Using deep reinforcement learning, we were able to reduce maintenance spending by approximately 50% compared to optimized, time-based maintenance strategies for the chosen system. A key component of our methodology is the integration of Shapley values to quantify the contribution of various factors to the decision-making process. This addition enhances the explainability and trustworthiness of our decisions, providing operators with transparent and understandable insights into the rationale behind maintenance strategies. The robustness and resiliency of our decision policy against observation noise were also thoroughly evaluated, demonstrating its effectiveness in uncertain operational environments.

Gaussian-inverse gamma mixture distribution–based extended Kalman observer for dynamic positioning control system of offshore platform

Under different working conditions and control objectives, fuzzy model predictive control is used to optimize the control effect by adjusting the weight to achieve stable and accurate control for the dynamic positioning control of offshore platforms. Also, aiming at the situation where the measurement noise of dynamic positioning control may be time-varying and non-Gaussian and the statistics of the noise are not exact, this paper proposes a variational Bayesian extended Kalman filter observer. In order to model non-Gaussian noise more reasonably, the joint posterior distribution of non-Gaussian noise with state and time-varying covariance is described as the product of Gaussian and independent inverse gamma distributions. Then, the variational Bayesian (VB) method was used to simultaneously estimate the system state, the intermediate variable values of the new distribution, and the noise parameters through fixed-point iteration. The fuzzy algorithm is applied to the model predictive control, and the weight parameters of the model predictive control are adaptively changed to better adapt to the changing working conditions in the control process. The simulation results show that the variational Bayesian extended Kalman filter observer has better performance than the existing algorithms when dealing with the time-varying non-Gaussian observation noise, whose statistics are not exact. The variational Bayesian extended Kalman filter observer can significantly improve the accuracy of the control when it is used in fuzzy model predictive control.

Robust recursive sigma point Kalman filtering for Huber-based generalized M-estimation

For nonlinear state estimation driven by non-Gaussian noise, the estimator is required to be updated iteratively. Since the iterative update approximates a linear process, it fails to capture the nonlinearity of observation models, and this further degrades filtering accuracy and consistency. Given the flaws of nonlinear iteration, this work incorporates a recursive strategy into generalized M-estimation rather than the iterative strategy. The proposed algorithm extends nonlinear recursion to nonlinear systems using the statistical linear regression method. The recursion allows for the gradual release of observation information and consequently enables the update to proceed along the nonlinear direction. Considering the correlated state and observation noise induced by recursions, a separately reweighting strategy is adopted to build a robust nonlinear system. Analogous to the nonlinear recursion, a robust nonlinear recursive update strategy is proposed, where the associated covariances and the observation noise statistics are updated recursively to ensure the consistency of observation noise statistics, thereby completing the nonlinear solution of the robust system. Compared with the iterative update strategies under non-Gaussian observation noise, the recursive update strategy can facilitate the estimator to achieve higher filtering accuracy, stronger robustness, and better consistency. Therefore, the proposed strategy is more suitable for the robust nonlinear filtering framework.

A Variational Formula of Forward-Backward Stochastic Differential System of Mean-Field Type with Observation Noise and Some Application

A Variational Formula of Forward-Backward Stochastic Differential System of Mean-Field Type with Observation Noise and Some Application

Path planning of 6-DOF free-floating space robotic manipulators using reinforcement learning

This paper presents a study on path planning for 6-DOF free-floating space robotic manipulators using Deep Deterministic Policy Gradient-based Reinforcement Learning. The focus is the development of a novel reward function tailored to address critical requirements for efficient and effective manipulation in space. These requirements include accurate pose alignment between the end-effector and the target, collision avoidance with both the target and other links of the manipulator, smoothing of joint velocities, adaptability to strong dynamic coupling between the manipulator and its base spacecraft due to high manipulator-spacecraft mass ratio, and resilience to noise in the state observations. Uniquely, the proposed reward function employs quaternions for orientation control to reduce pose misalignments and dynamic singularities, as opposed to traditional Euler angles. Our findings demonstrate that the Reinforcement Learning algorithm, when guided by this new reward function that integrates these enhancements and constraints, not only achieves the desired path planning objectives more efficiently but also exhibits faster convergence. Furthermore, the Reinforcement Learning successfully manages significant dynamic coupling effects caused by a high mass ratio between the robotic manipulator and the base spacecraft. Even under the challenge of noisy state observations, the trained agent successfully completes the path planning task, proving the Reinforcement Learning's applicability to real-space mission designs where the noise in observation is inevitable. The study highlights the critical role of reward function design in the Reinforcement Learning training process and its consequential impact on the solution quality, providing a solid foundation for future advancements in free-floating space robotic missions.

Intrinsic Rewards for Exploration Without Harm From Observational Noise: A Simulation Study Based on the Free Energy Principle.

In reinforcement learning (RL), artificial agents are trained to maximize numerical rewards by performing tasks. Exploration is essential in RL because agents must discover information before exploiting it. Two rewards encouraging efficient exploration are the entropy of action policy and curiosity for information gain. Entropy is well established in the literature, promoting randomized action selection. Curiosity is defined in a broad variety of ways in literature, promoting discovery of novel experiences. One example, prediction error curiosity, rewards agents for discovering observations they cannot accurately predict. However, such agents may be distracted by unpredictable observational noises known as curiosity traps. Based on the free energy principle (FEP), this letter proposes hidden state curiosity, which rewards agents by the KL divergence between the predictive prior and posterior probabilities of latent variables. We trained six types of agents to navigate mazes: baseline agents without rewards for entropy or curiosity and agents rewarded for entropy and/or either prediction error curiosity or hidden state curiosity. We find that entropy and curiosity result in efficient exploration, especially both employed together. Notably, agents with hidden state curiosity demonstrate resilience against curiosity traps, which hinder agents with prediction error curiosity. This suggests implementing the FEP that may enhance the robustness and generalization of RL models, potentially aligning the learning processes of artificial and biological agents.

Online numerical simulation method based on adaptive unscented Kalman filter for shear wall structures

Hybrid test method is an effective approach to evaluate the seismic performance of engineering structures. Particularly, online numerical simulation method (ONSM) based on unscented Kalman filter (UKF) model updating significantly improves the accuracy of the hybrid test results. However, the sensitivity of UKF to the initial values of the state vector, state covariance, and observation noise covariance may lead to the distortion of hybrid test results. To address these issues, one novel ONSM based on adaptive UKF (ONSM-AUKF) is proposed in this paper. The AUKF algorithm incorporates an adaptive variance module to maintain equilibrium. This not only reduces the sensitivity of UKF to the initial parameter settings, prevents filter divergence, but also enhances the accuracy and stability of estimation. The proposed ONSM-AUKF accurately identifies the constitutive model parameters by utilizing measured data from specimens through the AUKF algorithm, which improves the accuracy and the stability of parameter estimation. The simulation results indicate that AUKF algorithm has the capability to decrease the sensitivity of initial parameter settings, thus enhancing the stability and robustness of parameter estimation in comparison to the UKF algorithm. The experimental results validate the feasibility and effectiveness of the proposed ONSM-AUKF, as well as the advantages of the ONSM-AUKF over the ONSM-UKF. The results indicate that the proposed ONSM-AUKF may have broad application prospect in evaluating the seismic performance of various of complex engineering structures.

Multi-frequency smartphone positioning performance evaluation: insights into A-GNSS PPP-B2b services and beyond

In August 2023, Xiaomi unveiled the Redmi K60 Ultra, the first multi-frequency smartphone integrated with BeiDou-3 Navigation Satellite System Precise Point Positioning (PPP-B2b) services and employing PPP technology as the primary positioning method. The positioning enhancement service is provided by the Assisted Global Navigation Satellite System (A-GNSS) location platform developed by the China Academy of Information and Communications Technology. The signaling interaction between the server and the users strictly adheres to the Third Generation of Mobile Communications Technology Partnership Project Long-Term Evolution Positioning Protocol and the Open Mobile Alliance Secure User Plane Location framework. To comprehensively evaluate the Redmi K60 Ultra’s capabilities, this study designed six distinct experimental scenarios and conducted comprehensive research on multi-frequency and multi-GNSS observation noise, Time to First Fix (TTFF), as well as the performance of both GNSS-based and network-based positioning. Experimental results indicate that the GNSS chipset within the Redmi K60 Ultra has achieved a leading position in the consumer market concerning supported satellite constellations, frequencies, and observation accuracy, and is comparable to some low-cost GNSS receivers. A-GNSS positioning can reduce the TTFF from 30 to under 5 s, representing an improvement of over 85% in the cold start speed compared to a standalone GNSS mode. The positioning results show that the A-GNSS PPP-B2b service can achieve positioning performance with RMS errors of less than 1.5 m, 2.5 m, and 4 m in open-sky, realistic, and challenging urban environments. Compared to GNSS-based positioning, cellular network-based Observed Time Difference of Arrival (OTDOA) positioning achieves an accuracy ranging from tens to hundreds of meters in various experimental scenarios and currently functions primarily as coarse location determination. Additionally, this study explores the potential of the Three-Dimensional Mapping-Aided (3DMA) GNSS algorithm in detecting Non-Line-of-Sight signals and enhancing positioning performance. The results indicate that 3DMA PPP, as compared to conventional PPP, can significantly accelerate PPP convergence and improve positioning accuracy by over 30%. Consequently, 3D city models can be utilized as future assistance data for the A-GNSS location platform.

Computational phenotyping of aberrant belief updating in individuals with schizotypal traits and schizophrenia

BackgroundPsychotic experiences are thought to emerge from various interrelated patterns of disrupted belief updating, such as overestimating the reliability of sensory information and misjudging task volatility. Yet, these substrates have never been jointly addressed under one computational framework and it is not clear to what degree they reflect trait-like computational patterns. MethodsWe introduced a novel hierarchical Bayesian model that describes how individuals simultaneously update their beliefs about the task volatility and noise in observation. We applied this model to data from a modified Predictive inference task in a test-retest study with healthy volunteers (N=45, 4 sessions) and examined the relationship between model parameters and schizotypal traits in a larger online sample (N = 437) and in a cohort of patients with schizophrenia (N = 100). ResultsThe interclass correlations were moderate to high for model parameters and excellent for averaged belief trajectories and precision-weighted learning rates estimated through hierarchical Bayesian inference. We found that uncertainty about the task volatility was related to schizotypal traits and to positive symptoms in patients, when learning to gain rewards. In contrast, negative symptoms in patients were associated with more rigid beliefs about observational noise, when learning to avoid losses. ConclusionThese findings suggest that individuals with schizotypal traits across the psychosis continuum are less likely to learn or utilize higher-order statistical regularities of the environment and showcase the potential of clinically relevant computational phenotypes for differentiating symptom groups in a transdiagnostic manner.