Search Results (154)

Background/Objectives: Diagnostic laboratory tests are often judged by sensitivity and specificity, together with related measures of analytic validity. These assist in deciding on which tests are expected to be more accurate or precise, but they do not determine whether a test result improves patient care. A laboratory test has clinical value when it changes a provider’s decision in a way that improves expected outcomes after accounting for uncertainty and patient burden. This work presents DECIDE-Lab (Decision Centered Evaluation of Clinical Information and Dynamic Evidence for Laboratory Testing), an integrative decision-theoretic framework that adapts value-of-information (VOI) and POMDP methods to laboratory test selection to evaluate the value of single tests and serial testing. Methods: The framework connects test performance attributes to clinical utility through posterior belief updating and action thresholds. The concept of clinical utility frontiers is presented, which identifies dominated lab tests. DECIDE-Lab allows preference-sensitive and equity-aware evaluation of lab test diagnostic value by incorporating variation in test performance across patient subgroups. Results: An illustrative acute coronary syndrome application shows that the value of serial troponin testing concentrates in intermediate-risk patients for whom an additional result can change disposition or treatment. A worked sepsis pathway demonstrates the POMDP mechanics step by step from an initial belief state to the resulting stop-or-continue decision. Additional disease-state applications demonstrate how the framework can support diagnostic stewardship and adaptive ordering. Conclusions: The central implication is that diagnostic value should be measured by action-changing utility rather than accuracy alone, while examining how conclusions change with utilities and implementation constraints. Full article

Background: The analytic thinking score (ATS) has been interpreted as a linguistic marker of organized thought. Written language typically shows a higher ATS than spoken language. From the perspective of active inference under the free-energy principle, we proposed a preliminary neurocomputational model of this writing-over-speaking advantage. We propose that ATS is an externally computed linguistic measure reflecting analytic-thinking active states that arise from precision-weighted inference over internal conceptual-organization (CO) states. We hypothesize that written production shows a higher ATS when the active-inference agent increases posterior confidence in high-CO states. Methods: ATSs were extracted from written and spoken samples produced by university students who described thematic apperception test images. Participants were modeled as active-inference agents using a two-timestep Markov decision process (MDP) in which speaking and writing sensory cues updated beliefs about internal CO states which then drove analytic thinking active states. Belief updating was formalized through marginal message-passing and theoretically interpreted in terms of prediction-error signaling and precision-weighted neuronal synaptic gain. An attention-related parameter (AP) controlled the precision of the CO-sensory state mapping. Bayesian model selection was used to assess the model’s preliminary construct validity. Results: Written responses showed higher ATSs than spoken responses. The AP estimate indicated that writing cues supported posterior inference toward high-CO states stronger than speaking cues. Bayesian model selection favored the active-inference MDP over a Variational Laplace linear model. Conclusions: The current preliminary evidence speaks to a candidate active-inference model in which writing ascribes higher precision-weighted inference of CO, reflected in higher ATSs. Full article

Advanced Persistent Threats (APTs) pose a serious threat to Internet of Things (IoT) systems because of their stealthiness, persistence, and ability to adapt to defensive responses. Communication-Based Train Control (CBTC) systems, as IoT-enabled railway signaling infrastructures, have evolved from relatively closed operational environments into interconnected cyber-physical networks, exposing train control systems to coupled cyber intrusion and operational-safety risks. To address this challenge, this paper proposes an alert-driven active defense framework for CBTC systems that integrates Bayesian belief updating, hypergame-based cognitive-bias modeling, and Hierarchical Reinforcement Learning (HRL). The framework converts intrusion detection system (IDS) alerts, network traffic observations, and cyber-physical observations into belief-state, transition, and reward inputs. The Bayesian model estimates attacker type and attack stage, the hypergame model represents deception-induced asymmetric cognition between attackers and defenders, and the HRL decouples strategic defense posture selection from tactical defense execution. The scenario-driven simulations in a CBTC APT defense setting show that the proposed model strategy achieves an 87.1% defense success rate against APT attacks while consuming 62.7% of the normalized defense resources, outperforming DQN, PG, and PPO under the same test conditions. These results suggest that explicitly coupling cyber observations, CBTC operational constraints, and hierarchical deception-aware policies can improve cost-aware active defense for railway signaling infrastructures. Full article

Large vision–language models have advanced embodied navigation by integrating visual perception with natural-language reasoning. However, vision-and-language navigation (VLN) for unmanned aerial vehicles in low-altitude urban airspaces remains challenging due to occluded views, dynamic layouts, limited communication bandwidth, and partial observability. Existing methods mainly focus on single-agent egocentric navigation and lack explicit modeling of uncertainty and inter-agent dependencies in collaborative multi-UAV settings. We propose Collaborative Low-Altitude Space Navigation (Co-LASN), a dynamic Bayesian network-based framework for collaborative VLN in embodied multi-agent systems. Co-LASN jointly models environmental dynamics, linguistic constraints, and inter-agent dependencies in a unified probabilistic representation, allowing each UAV to update its belief state and incorporate information from neighboring agents when making navigation decisions. Experiments on a low-altitude subset of the HaL-13k benchmark show that, under the evaluated simulation protocol, Co-LASN achieves higher navigation metrics than single-agent and partially collaborative baselines. In the 3-agent setting, Co-LASN increases the any-success rate (ASR) from $12.37\%$ to $15.23\%$ and reduces the min navigation error (MNE) from $99.86$ to $89.46$. These results demonstrate the relative effectiveness of belief-aware collaboration within the evaluated simulation setting. Full article

Accurate prediction of multivariate effluent quality is essential for achieving reliable operation and sustainable management of wastewater treatment processes (WWTPs). However, the strong nonlinearity, coupling relationships, and non-prioritized multi-input multi-output (MIMO) characteristics of WWTP pose significant challenges to conventional prediction methods. To address these issues, a vectorized evidential reasoning-based multivariate effluent quality (VER-MEQ) prediction method is proposed. First, a VER model is developed, in which the nonlinear mapping between multiple process variables and multiple effluent quality indicators is established through a vector evidence matrix (VEM), enabling simultaneous online prediction of multiple outputs within a unified inference framework. Subsequently, a structured hybrid initialization (SHI) strategy is introduced to improve the initialization quality of the genetic algorithm, and the VER inference process is incorporated into parameter optimization to enable online model parameter updating, thereby improving prediction performance. The proposed method is validated under sunny, rainy, and stormy operating scenarios. Experimental results demonstrate that VER-MEQ achieves competitive prediction accuracy, provides a transparent belief-based inference process, and maintains preliminary anti-interference performance under the tested conditions. By providing transparent and credible prediction results for effluent ammonia nitrogen (NH₃-N_e) and total nitrogen (TN_e), the proposed framework can support proactive operational decision-making, improve effluent compliance, reduce the risk of nutrient discharge, and contribute to the sustainable operation of WWTPs. Full article

In online retail, consumers cannot experience product quality before purchase. With the adoption of artificial intelligence (AI), platforms can certify product quality information. However, stronger platform certification may reduce sellers’ incentives to disclose and limit personalized information such as product fit. This study examines the conditions under which a platform should adopt AI-assisted platform certification (AIPC). We develop a game-theoretic model with one platform and two competing sellers. We compare the case of not adopting AIPC with adopting AIPC, and examine how AIPC affects seller disclosure, pricing, and profits. Sellers decide whether to disclose product information and set prices. Consumers update their quality beliefs based on seller disclosure and platform labels. Our results show that AIPC is not always the preferred strategy. When product-fit information spillovers between competing sellers are strong, the platform may be better off not adopting AIPC. When information spillovers are weak, AIPC adoption depends on consumers’ prior belief regarding product quality. Specifically, when consumers have a low prior belief that an uncertified or undisclosed product is of high quality, AIPC benefits the platform and sellers but reduces consumer surplus. When this prior belief is sufficiently high, AIPC creates a win–win–win outcome for the platform, sellers, and consumers. Full article

This paper presents a computational model of delirium in the Intensive Care Unit (ICU), in which delirium is defined as the endpoint of a self-reinforcing cycle of predictive failure between two bidirectionally coupled agents: the patient and the ICU room environment. Drawing on the active inference framework and the free energy principle, the paper proposes that delirium is not a property of the patient in isolation but a relational phenomenon that emerges when the environment persistently fails to predict the patient’s internal state. This failure triggers a causal feedback mechanism in which desynchronization pressure progressively sharpens the patient’s prior beliefs—implementing precision rigidity in the correct active inference sense: not a brain overwhelmed by noise but a brain locked into a state that incoming observations can no longer update. The model is implemented as a two-agent POMDP in which both agents maintain generative models and continuously attempt to predict each other’s states. The room agent (R)—understood as the environment-side sensing–inference–actuation loop, whether instantiated by clinical staff or by an automated monitoring system—infers the patient (P)’s latent parameters $({\theta }_{\mathrm{cog}},{\theta }_{\mathrm{emo}})$ over time and builds a progressively personalized generative model of the patient. Synchronization is operationalized via two commensurable directional surprisal metrics: $S_{R\to P}=-ln{Q}_R\left(s^{*}\right)$, the room’s surprisal at the patient’s true state, and $S_{P\to R}=-lnP(o_R\mid Q_P)$, the patient’s surprisal at the room’s observations. A systematic ablation study across four model variants shows that room inference is the architectural component necessary to reproduce the synchronization–delirium relationship: when the room infers, the association between synchronization and declared delirium is strong and stable, whereas a non-inferring room collapses to ceiling delirium rates and a weak association. $\theta$ learning and the prior-sharpening feedback do not increase the strength of this association; instead they shape the phenotypic gradient, reducing ceiling effects in vulnerable phenotypes and amplifying the separation between them. The model is presented as a computational hypothesis generator rather than a calibrated clinical predictor, and its implications for ICU design are discussed. Full article

The transition towards sustainable manufacturing necessitates complex optimization that integrates economic goals with environmental factors, such as energy consumption and greenhouse gas emissions. This research addresses the critical challenge of optimizing the Incoming Quality Control (IQC) policy for raw material batches. The primary objective is formulated as a multi-criteria control problem that jointly minimizes the weekly final product cost, carbon footprint, and energy consumption. To handle sequential decision making under uncertainty, we adopt a scalarized reinforcement learning (RL) reward that combines these objectives into a single value function and explores different trade-offs through alternative weight configurations. To effectively handle the uncertainty in incoming quality and the sequential decision making required for dynamic control, the optimization problem is modeled as a Bayesian Adaptive Markov Decision Process (BAMDP). To maintain computational tractability despite the continuous belief space inherent in the BAMDP formulation, we employ a Deep Q-Network (DQN) architecture acting as an approximate dynamic programming solver. The Bayesian framework represents model uncertainty explicitly, updates beliefs as new inspection evidence becomes available, and allows prior domain knowledge on supplier quality to be incorporated into the learning process. The BAMDP formulation is used to learn a set of adaptive inspection policies that adjust the IQC strategy over time to achieve conflicting goals: reducing inspection costs while maintaining standard quality, minimizing energy consumption, and lowering CO₂-equivalent emissions. The goal is to find robust policies that balance these trade-offs under different quality and demand conditions. This methodology aligns with the principles of Industry 5.0 by leveraging advanced artificial intelligence (AI) methods, such as reinforcement learning (RL), coupled with a stochastic simulation of the production system, based on a geometric/physical model of the component’s tolerance chains, to support decision-makers in designing and assessing sustainable IQC strategies. Comparative simulations on the case study, including a benchmark against ISO 2859-1 sampling plans, confirm that this dynamic and risk-aware optimization paradigm can reduce overall cost, energy use, and environmental impact across various quality conditions, while preserving outgoing quality. Full article

Power grid technical renovation projects are implemented through project-based supply chains involving equipment procurement, logistics coordination and on-site construction under market, delivery and carbon constraints. Their final cost is jointly affected by engineering quantities, supplier behavior, lead-time uncertainty, material price volatility and sustainability requirements. Existing studies usually emphasize technical parameters and direct expenditure, whereas supplier reliability, green procurement, carbon intensity and procurement contingency effects are only indirectly incorporated. This study develops a dynamic Bayesian model for carbon-adjusted cost forecasting and investment priority support in power grid technical renovation projects. Based on 800 anonymized project-level records, a random forest is first used to identify informative engineering, supply chain and sustainability variables. These variables are then organized in a Bayesian network that links observed evidence, intermediate cost nodes and the carbon-adjusted cost target. A dynamic evidence-weighting mechanism updates posterior cost beliefs as supplier, logistics, market and carbon information become available during implementation. Compared with static Bayesian inference, XGBoost, an improved BPNN and GRA-based benchmarks, the proposed model yields lower MAE and RMSE. Ablation and scenario analyses further show that supply chain and sustainability variables improve both predictive performance and decision interpretability. The results provide a quantitative basis for budget control, green procurement adjustment, contingency allocation and sustainable asset renewal prioritization in energy enterprises. Full article

Bayesian inference is predominantly formulated in a continuous framework, in which posterior beliefs are represented by smooth probability densities. However, an alternative discrete representation—already implicit in Bayes’s original construction—remains conceptually distinct and structurally informative. This paper develops a representation-level analysis of Bayesian updating in the binomial setting and shows that discrete and continuous posteriors may exhibit qualitatively distinct behavior under finite parameter resolution. In particular, coarse discretization can induce regime-dependent divergence from the continuous posterior, even when the algebraic form of the likelihood is identical. The analysis further demonstrates that divergence is not determined solely by grid resolution but also by the balance between prior strength and sample size. By introducing a scale-dependent perspective in which representational resolution and prior magnitude jointly define distinct regimes of inference, the paper clarifies how structural and analytic descriptions interact under finite conditions. Full article

Real-world autonomous agents learn under nonstationarity, safety constraints, and finite energetic budgets. We develop a framework for perennial learning—agents that continuously refine their models while provably controlling the cost of forgetting—by unifying three classical pillars: Kolmogorov complexity, which equates scientific discovery with algorithmic compression; Landauer’s principle, which assigns a minimal thermodynamic cost of $k_{\mathrm{B}}Tln2$ per erased bit to every irreversible model update; and port-Hamiltonian (PH) dynamics, whose $(J-R)\nabla H$ decomposition separates zero-cost reversible inference from costly irreversible forgetting by construction. The Maxwell demon analogy is formalized: each learning episode is a Szilard cycle in which information acquisition, belief transport, and memory erasure must balance thermodynamically. The information-distance framework, comprising the normalized information distance (NID) and normalized compression distance (NCD), provides a computable geometry for measuring learning progress and guiding curriculum design. We separate theideal uncomputable regularizer based on prefix complexity from the practical compressor/MDL (minimum description length) surrogate that appears in optimization and prove a calibration lemma linking the two under a mild uniform-accuracy assumption. Under explicit regularity, compact-sublevel, and non-energy-extracting assumptions, we prove a passivity speed limit for curriculum-induced contractions of the effective feasible set. Under local asymptotic normality, we reprove that Fisher information is a local posterior codelength proxy rather than an exact theorem about algorithmic entropy. A conditional sequential information-budget proposition shows that the per-stage sample requirement scales as $\tilde{O}(\Delta k_t/{\lambda }_{\star })$, where $\Delta k_t$ is the number of materially changed model coordinates (not the total model complexity $k_t$); the $k^3\to \Delta k$ improvement is conditional on a warm-start assumption and a chosen cold-start baseline. A double-integrator running example with a moving obstacle illustrates the architecture. Full article

Sense of agency (SoA) refers to the experience or belief that one’s own actions directly lead to a particular event in the body or external world. SoA does not arise from a single mechanism; rather, evidence suggests the integration of predictive, sensory, and contextual cues whose relative weighting may vary across tasks and individuals. Since this process requires the maintenance, comparison, and updating of action-related information, working memory-related processes have been proposed as a candidate contributor to selected components of agency. At the same time, aging is associated with alterations in temporal processing, sensory precision, executive control, and other functions that may change the use of agency-related cues. However, the extent to which age-related differences in SoA can be explained by working memory-related processes remains undefined. This narrative review examines the theoretical and empirical basis for considering working memory as a candidate mechanism linking aging and SoA. Specifically, it discusses major theoretical frameworks of agency, measurement approaches, evidence on aging-related differences in SoA, and studies suggesting that cognitive load- and working memory-related factors may modulate agency-related processing. Current evidence is suggestive, although it remains indirect, highlighting key methodological limitations and outlining priorities for future research to clarify how aging- and working-memory-related processes may shape different components of agency. Full article

Multi-unmanned aerial vehicle (UAV) platforms integrate radio-frequency (RF) sensing, datalinks, and onboard embedded compute; adversarial electronic warfare (EW) degrades these subsystems through jamming and forces decentralized control policies to act on fragmented observations—a setting aligned with intelligent electronic systems and autonomous robotics in contested spectrum. Cooperative swarms then face two compounding failure modes: loss of coherent situational awareness, and reward-driven passive survival that suppresses mission completion. Memory-based multi-agent reinforcement learning (MARL) partially addresses the first but tends to reinforce the second; dense intent shaping addresses the second but becomes unreliable when observations are incomplete. We propose CIDA (Cognitive–Intent Dual-Stream Architecture), a reinforcement learning framework that decouples belief reconstruction from tactical intent at the representation level while coupling them through a unified actor–critic update. The cognitive stream encodes a 64-step observation history with a pre-normalized Transformer to reconstruct threat belief; the intent stream supplies a hierarchical potential field (reconnaissance, threat-weighted engagement, and approach incentives). A steady-state training mechanism (dynamic reward scaling and adaptive gradient clipping) stabilizes Transformer-based on-policy learning under non-stationary multi-agent dynamics. In a complex terrain scenario with SAM, AAA, and jammer assets, CIDA reaches 96.15% task success versus 12.21% (memoryless PPO) and 25.28% (MAPPO+RNN), with ablations showing nonlinear coupling and emergent tactics such as jammer bypass and weak-sector traversal. Results are robust to a four-fold sweep of the intent-shaping weight (above 90% success). Full article

Multi-UAV cooperative search is important for applications such as target reconnaissance, environmental monitoring, and emergency response. In practice, communication is often spatially heterogeneous due to terrain occlusion and environmental interference, which may delay information sharing and weaken coordination efficiency when UAVs traverse communication-blocked areas. To address this issue, we propose an Age of Information (AoI)- and Dempster–Shafer (DS)-enhanced cooperative search framework for multi-UAV systems under spatially structured communication constraints. Specifically, a DS belief map is introduced to fuse uncertain observations, while AoI is used to characterize the freshness of delayed information. An AoI-aware update mechanism further integrates buffered observations into the global belief map after communication recovery. The search process is then formulated as a communication-aware multi-agent sequential decision-making problem and solved using reinforcement learning. To demonstrate the generality of the proposed framework, we instantiate it with Proximal Policy Optimization (PPO), Multi-Agent Proximal Policy Optimization (MAPPO), and Q-value Mixing Network (QMIX). Experimental results show that the proposed framework consistently outperforms the baseline methods under heterogeneous environments and different communication conditions. Among all variants, AoI-DS-MAPPO achieves the best overall performance, improving average reward, success rate, and the number of detected targets by 26.13%, 24.32%, and 3.65%, respectively, while reducing episode length by 31.96% relative to the strongest baseline. Full article

Aiming to improve radar multi-target tracking (MTT) accuracy and association performance in complex scenarios involving dense clutter, missed detections, and maneuvering targets, an improved tracklet generation approach based on the expectation–maximization (EM) framework is proposed in which data association variables and motion model variables are jointly modeled as latent variables. These variables are estimated through iterative updates based on the loopy belief propagation (LBP) algorithm and the interacting multiple model (IMM) filtering and smoothing algorithms to generate high-confidence tracklets. Then, a delayed decision-making strategy based on the multi-hypothesis approach is employed to associate these tracklets into complete target trajectories. The resulting algorithm is named IMM-TrackletMHT. The performance of the IMM-TrackletMHT algorithm is evaluated and compared with several baseline algorithms in simulated scenarios under different clutter rates and detection probabilities. The simulation results demonstrate that the proposed algorithm consistently outperforms the baseline methods in terms of tracking accuracy, exhibits strong robustness to variations in the operating environment, and achieves higher computational efficiency in multi-scan measurement processing, thereby demonstrating the effectiveness and superiority of the proposed tracklet generation approach for maneuvering MTT. Full article