Predictive Processing Research Articles

Combining predictive soil mapping and process models to estimate future carbon sequestration potential under no-till

There is increasing interest in soil organic carbon (SOC) sequestration as a climate change mitigation strategy. There is a need to estimate the quantity of SOC sequestered historically due to no-till, and the remaining sequestration potential in Saskatchewan. To answer this question, predictive soil mapping results were linked with the Century model to predict SOC stock change over time to a depth of 20 cm considering three different future climate change scenarios. Climate scenarios included low, moderate, and high amounts of climate change and included estimated changes to monthly minimum, average, and maximum temperature, total monthly precipitation, and average monthly relative humidity at an 800 m × 800 m resolution. Historically, the modelled average SOC gain for Saskatchewan was 2.8 Mg ha−1. Future potential simulated SOC was lower over the next 30 years, with average SOC gains estimated to range from 1.4 to 1.7 Mg ha−1 by 2054 and 2.3 to 3.1 Mg ha−1 by 2100. There is also unequal spatial distribution of SOC stock gain potential, with the northern grain growing regions showing lower future potential. The predicted future gains will be at a lower rate than in the past with carbon sequestration rates dropping from 0.06 to less than 0.02 Mg ha−1 year−1. Additional management practices such as improved residue management and the introduction of crop varieties with increased below ground carbon inputs and more stable residues should be explored to offset the diminishing SOC returns from no-till.

Anytime valid and asymptotically efficient inference driven by predictive recursion

Abstract Distinguishing two candidate models is a fundamental and practically important statistical problem. Error rate control is crucial to the testing logic but, in complex nonparametric settings, can be difficult to achieve, especially when the stopping rule that determines the data collection process is not available. This paper proposes an e-process construction based on the predictive recursion algorithm originally designed to recursively fit nonparametric mixture models. The resulting predictive recursion process affords anytime valid inference and is asymptotically efficient in the sense that its growth rate is first-order optimal relative to the predictive recursion’s mixture model.

Oscillatory traveling waves provide evidence for predictive coding abnormalities in schizophrenia

BackgroundThe computational mechanisms underlying psychiatric disorders are hotly debated. One hypothesis, grounded in the Bayesian predictive coding framework, proposes that schizophrenia patients have abnormalities in encoding prior beliefs about the environment, resulting in abnormal sensory inference, which can explain core aspects of the psychopathology, such as some of its symptoms. MethodsHere, we tested this hypothesis by identifying oscillatory traveling waves as neural signatures of predictive coding. By analyzing an EEG dataset comprising 146 schizophrenia patients and 96 age-matched healthy controls, during resting states and a visual backward masking task. ResultsWe found that schizophrenia patients have stronger top-down alpha-band traveling waves compared to healthy controls during resting state, supposedly reflecting overly precise priors at higher levels of the predictive processing hierarchy. We also found stronger bottom-up alpha-band waves in schizophrenia patients during a visual task, in line with the notion of enhanced signaling of sensory precision errors. ConclusionsOur results yield a novel spatial-based characterization of oscillatory dynamics in schizophrenia, considering brain rhythms as traveling waves and providing a unique framework to study the different components involved in a predictive coding scheme. Altogether, our findings significantly advance our understanding of the mechanisms involved in fundamental pathophysiological aspects of schizophrenia, promoting a more comprehensive and hypothesis-driven approach to psychiatric disorders.

Predicting schedule adherence of engineering changes – a case study on effectivity date adherence prediction using machine learning

Engineering changes (EC), redesigns of components, are common with complex products. Their realisation into production systems is a lengthy process and thorough control is imperative. As with most complex processes, however, this is a resource-intensive task and support is required to direct scarce resources. In other domains, predictive business process monitoring (PBPM) research has demonstrated that complex processes can be monitored using machine learning approaches. However, it is indicated that predictive performance is domain sensitive. Therefore, we investigate and show how PBPM can be applied to the EC process by predicting adherence to planned implementation dates. As the outcome of a case study comparing 30 predictive models, our research indicates that EC effectivity date adherence prediction, and thus pre-emptive EC schedule monitoring, is possible. However, performance comparison hints that shallow learning algorithms outperform deep learning algorithms. Furthermore, as the optimal algorithm depends on the deployment scenario, it is demonstrated how cost curves are better decision criteria for choosing models compared to threshold dependent metrics. Based on these findings, this article offers a blueprint for developing machine learning models for predicting the EC schedule adherence and lays the base for further research towards automatic EC scheduling and control.

Processing Fluency and Predictive Processing: How the Predictive Mind Becomes Aware of its Cognitive Limitations

AbstractPredictive processing is an influential theoretical framework for understanding human and animal cognition. In the context of predictive processing, learning is often reduced to optimizing the parameters of a generative model with a predefined structure. This is known as Bayesian parameter learning. However, to provide a comprehensive account of learning, one must also explain how the brain learns the structure of its generative model. This second kind of learning is known as structure learning. Structure learning would involve true structural changes in generative models. The purpose of the current paper is to describe the processes involved upstream of these structural changes. To do this, we first highlight the remarkable compatibility between predictive processing and the processing fluency theory. More precisely, we argue that predictive processing is able to account for all the main theoretical constructs associated with the notion of processing fluency (i.e., the fluency heuristic, naïve theory, the discrepancy‐attribution hypothesis, absolute fluency, expected fluency, and relative fluency). We then use this predictive processing account of processing fluency to show how the brain could infer whether it needs a structural change for learning the causal regularities at play in the environment. Finally, we speculate on how this inference might indirectly trigger structural changes when necessary.

Evaluating a predictive model of avoidable hospital events for race- and sex-based bias.

To evaluate whether race- and sex-based biases are present in a predictive model of avoidable hospital (AH) events. We examined whether Medicare fee-for-service (FFS) beneficiaries in Maryland with similar risk scores differed in true AH event risk on the basis of race or sex (n = 324,834). This was operationalized as a logistic regression of true AH events on race or sex with fixed effects for risk score percentile. Beneficiary-level risk scores were derived from 36 months of Medicare FFS claims (April 2019-March 2022) and generated in May 2022. True AH events were observed in claims from June 2022. Black patients had higher average risk scores than White patients; however, the likelihood of experiencing an AH event did not differ by race when controlling for predicted risk (Marginal Effect [ME] = 0.0003, 95%CI -0.0003 to 0.0009). AH event likelihood was lower in males when controlling for risk level; however, the effect was small (ME = -0.0008, 95% CI -0.0013 to -0.0003) and it did not differ by sex for the target group for intervention (ME = 0.0002, 95% CI -0.0031 to 0.0036). We implemented a simple bias assessment methodology and found no evidence of meaningful race- or sex-based bias in this model. We encourage the incorporation of bias checks into predictive model development and monitoring processes.

Sex Trafficking Prevention for Youth with Intellectual and Developmental Disabilities: Service Provider Perspectives

ABSTRACT Given the paucity of research on sex-trafficking prevention education for youth with intellectual and developmental disabilities (IDD), this qualitative study aimed to determine service providers’ (N = 22) perspectives regarding (a) topics and delivery methods to include in this programming, (b) factors that promote accessibility, and (c) challenges and recommendations associated with implementation. Individual interviews and focus groups were conducted using Zoom. For programming acceptability, participants identified challenges (e.g. misconception that individuals with IDD are asexual) and recommendations (e.g. teaching developmentally appropriate concepts in early childhood). For content and delivery, topics included consent, safety, and context/social rules. Tactical activities for self-care and sensory processing, and non-infantilizing, realistic visuals were recommended to improve accessibility. For systems involvement in disclosures and referrals, participants noted challenges (i.e. confusion about agency responsibility and information sharing) and recommendations (i.e. ensuring a clear and predictable disclosure process for youth). Findings can contribute to policy development, particularly in service provider training for responding to disclosures and clarifying funding allocation for school and community sex trafficking prevention.

God-free energy and evolutionary epistemology

This article investigates the reasons behind interwovenness and the profound philosophical thought that a God must be a part of the brain’s predictive processing entropies. The article accomplishes this by utilising South African theologian Wentzel J. van Huyssteen’s epistemological consciousness as an added benefit in examining our planet and, as a result, our cosmic universe, which transcends human God notions. This article caters for academics interested in the science-religion dispute on the existence of life inside the integrated metaphysical God. In this article, the connection between metaphysicality, free energy and Van Huyssteen’s epistemic consciousness is used as metaphor(s) in our estimates of what life is, which includes a God, as it will show that we as humans can reasonably expect to exist beyond this (the) Earth. It is therefore the primary question that resonates with the whole of the universe as well as the minute fragmentation that is humans, only recently identified as sapiens, that sigh, or even better, long for a God, that encloses the question of: are we as humans made of, organs, metabolic systems, cells, atoms, memories or passions, that are a plausible mixture of an effective cognitive and affective cocktail, so to speak?Intradisciplinary and/or interdisciplinary implications: This article proposes a dual paradox that theology and science are not only in opposition yet are, juxtaposing (therefore integrating) one another. They share a very astute understanding of free energy from both a mathematical (natural sciences) and philosophical (theological) evolutionary epistemological perspective.

How does it end? Endpoints of boundaries lead to completion in macro-events.

While watching someone kicking a ball, missing moments of ball contact can be incorrectly identified as seen if the event is continued in a causal manner (i.e., the ball flying off). Does event completion also occur for events of a larger scale such as having breakfast (macro-event), which consists of multiple sub-steps like toasting bread (micro-event)? We conducted two experiments to measure event completion in macro-events presenting portions of multiple micro-events. In Experiment 1, video summaries were formed with or without event boundary information where a macro-end was either present or absent. Macro-end signified an overarching goal achievement that signaled the completion of previous tasks (such as leaving the kitchen with a full breakfast plate). More completion occurred for summaries with event boundary information and macro-ends. In Experiment 2, we tested two alternative hypotheses to explore the underlying process by showing the beginnings or ends of a micro-event. While the predictive processing hypothesis suggests that event completion is based more on predicting the future states of the event based on beginning information, the backward inferences hypothesis suggests that event completion relies more on deductions formed after the fact based on event endings. Results of Experiment 2 suggest that the ends of event boundaries lead to more event completion, possibly due to their role in forming causal connectivity. These results help to further understand event completion on a macro level.

How does depressive cognition develop? A state-dependent network model of predictive processing.

Depression is vastly heterogeneous in its symptoms, neuroimaging data, and treatment responses. As such, describing how it develops at the network level has been notoriously difficult. In an attempt to overcome this issue, a theoretical "negative prediction mechanism" is proposed. Here, eight key brain regions are connected in a transient, state-dependent, core network of pathological communication that could facilitate the development of depressive cognition. In the context of predictive processing, it is suggested that this mechanism is activated as a response to negative/adverse stimuli in the external and/or internal environment that exceed a vulnerable individual's capacity for cognitive appraisal. Specifically, repeated activation across this network is proposed to update individual's brain so that it increasingly predicts and reinforces negative experiences over time-pushing an individual at-risk for or suffering from depression deeper into mental illness. Within this, the negative prediction mechanism is poised to explain various aspects of prognostic outcome, describing how depression might ebb and flow over multiple timescales in a dynamically changing, complex environment. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Learned response dynamics reflect stimulus timing and encode temporal expectation violations in superficial layers of mouse V1.

The ability to recognize ordered event sequences is a fundamental component of sensory cognition and underlies the capacity to generate temporally specific expectations of future events based on previous experience. Various lines of evidence suggest that the primary visual cortex participates in some form of predictive processing, but many details remain ambiguous. Here we use two-photon calcium imaging in layer 2/3 of the mouse primary visual cortex (V1) to study changes to neural activity under a multi-day sequence learning paradigm with respect to prediction error responses, stimulus encoding, and time. We find increased neural activity at the time an expected, but omitted, stimulus would have occurred but no significant prediction error responses following an unexpected stimulus substitution. Sequence representations became sparser and less correlated with training, although these changes had no effect on decoding accuracy of stimulus identity or timing. Additionally, we find that experience modifies the temporal structure of stimulus responses to produce a bias towards predictive stimulus-locked activity. Finally, we find significant temporal structure during intersequence rest periods that was largely unchanged by training.

EchoPT: A Pretrained Transformer Architecture That Predicts 2D In-Air Sonar Images for Mobile Robotics.

The predictive brain hypothesis suggests that perception can be interpreted as the process of minimizing the error between predicted perception tokens generated via an internal world model and actual sensory input tokens. When implementing working examples of this hypothesis in the context of in-air sonar, significant difficulties arise due to the sparse nature of the reflection model that governs ultrasonic sensing. Despite these challenges, creating consistent world models using sonar data is crucial for implementing predictive processing of ultrasound data in robotics. In an effort to enable robust robot behavior using ultrasound as the sole exteroceptive sensor modality, this paper introduces EchoPT (Echo-Predicting Pretrained Transformer), a pretrained transformer architecture designed to predict 2D sonar images from previous sensory data and robot ego-motion information. We detail the transformer architecture that drives EchoPT and compare the performance of our model to several state-of-the-art techniques. In addition to presenting and evaluating our EchoPT model, we demonstrate the effectiveness of this predictive perception approach in two robotic tasks.

Diverse frontoparietal connectivity supports semantic prediction and integration in sentence comprehension.

Predictive processing in parietal, temporal, frontal, and sensory cortex allows us to anticipate future meanings to maximize the efficiency of language comprehension, with the temporoparietal junction (TPJ) and inferior frontal gyrus (IFG) thought to be situated towards the top of a predictive hierarchy. Although the regions underpinning this fundamental brain function are well-documented, it remains unclear how they interact to achieve efficient comprehension. To this end we recorded functional magnetic resonance imaging (fMRI) in 22 participants (11 males) while they comprehended sentences presented part-by-part, in which we manipulated the constraint provided by sentential contexts on upcoming semantic information. Using this paradigm, we examined the connectivity patterns of bilateral TPJ and IFG during anticipatory phases (i.e., before the onset of targets) and integration phases (i.e., after the onset of targets). When upcoming semantic content was highly predictable in strong-constraint contexts, both left TPJ and bilateral IFG showed stronger visual coupling, while right TPJ showed stronger connectivity with regions within control, default mode, and visual networks, including IFG, parahippocampal gyrus, posterior cingulate, and fusiform gyrus. These connectivity patterns were weaker when predicted semantic content appeared, in line with predictive coding theory. Conversely, for less predictable content, these connectivity patterns were stronger during the integration phase. Overall, these results suggest that both top-down semantic prediction and bottom-up integration during predictive processing are supported by flexible coupling of frontoparietal regions with control, memory, and sensory systems.Significance Statement Recent work has revealed the neural basis of predictive language comprehension. However, it remains unclear how brain regions change their connectivity in a dynamic fashion to support comprehension in highly predictive and less predictive contexts. Here, we show that stronger frontoparietal connectivity with cognitive control, memory, and sensory areas supports top-down prediction generation in strong-constraint contexts; these connectivity patterns are reduced when the anticipated information appears. This pattern is reversed when upcoming sensory input is unpredictable; connectivity is stronger after word inputs have been presented, allowing semantic integration with preceding low-constraint context. Our findings suggest that both top-down semantic prediction and bottom-up semantic integration in language comprehension rely upon diverse functional coupling of higher-order frontoparietal regions with other brain systems.

The Role Of Machine Learning In Transforming Healthcare: A Systematic Review

This systematic review explores the transformative role of machine learning (ML) in healthcare, focusing on its applications in disease diagnosis, personalized medicine, healthcare operations, and patient monitoring. By analyzing 167 peer-reviewed studies published between 2015 and 2024, the review highlights how ML algorithms have significantly improved early disease detection, treatment personalization, and operational efficiencies in healthcare settings. Findings reveal that ML-powered diagnostic tools, especially deep learning models, demonstrate accuracy levels comparable to human experts in areas like cancer detection. Additionally, ML models tailored for personalized medicine have shown promise in optimizing treatment protocols based on genetic profiles, reducing adverse reactions and improving patient outcomes. In healthcare operations, ML applications have streamlined hospital resource management, enhanced workflow efficiency, and reduced administrative burdens through predictive analytics and natural language processing (NLP). Moreover, ML-driven remote patient monitoring systems have enabled proactive interventions for chronic disease management. Despite these advancements, challenges related to data privacy, algorithmic transparency, and regulatory compliance persist. Addressing these barriers is critical for realizing the full potential of ML in healthcare, ensuring ethical deployment and equitable access. The review concludes by emphasizing the need for robust regulatory frameworks and further research to enhance the integration of ML technologies into clinical practice.

Large-Scale Low-Rank Gaussian Process Prediction with Support Points

Low-rank approximation is a popular strategy to tackle the “big n problem” associated with large-scale Gaussian process regressions. Basis functions for developing low-rank structures are crucial and should be carefully specified. Predictive processes simplify the problem by inducing basis functions with a covariance function and a set of knots. The existing literature suggests certain practical implementations of knot selection and covariance estimation; however, theoretical foundations explaining the influence of these two factors on predictive processes are lacking. In this article, the asymptotic prediction performance of the predictive process and Gaussian process predictions are derived and the impacts of the selected knots and estimated covariance are studied. The use of support points as knots, which best represent data locations, is advocated. Extensive simulation studies demonstrate the superiority of support points and verify our theoretical results. Real data of precipitation and ozone are used as examples, and the efficiency of our method over other widely used low-rank approximation methods is verified. Supplementary materials for this article are available online, including a standardized description of the materials available for reproducing the work.

Fermatean Fuzzy TOPSIS Method and Its Application in Ranking Business Intelligence-Based Strategies in Smart City Context

With the expansion of smart cities, the use of business intelligence (BI) has emerged as a crucial tool for resource optimization, increasing efficiency, and improving the citizens’ quality of life. BI enables companies to make better strategic decisions by analyzing vast amounts of urban data, helping them remain competitive in the dynamic smart city environment. This study utilizes content analysis and the Fermatean Fuzzy TOPSIS (FF-TOPSIS) method to rank the strategies based on business intelligence in the context of smart city. Initially, relevant criteria were identified through content analysis, and subsequently, five strategies were developed and ranked based on these criteria. The results revealed that the "Development of IOT-enabled smart networks (S2)" ranked highest, as it plays a significant role in optimizing resource management and enhancing urban service performance, thereby contributing greatly to the advancement of smart cities. "Process automation and the deployment of robotic systems (S5)" ranked second, as it enhances efficiency and reduces human errors. "Cloud platform integration for seamless access to data and services (S3)" also proved to be of considerable importance, ranking third, as it provides seamless access to data and services. " Artificial intelligence deployment for predictive analytics and process optimization (S4)" ranked fourth and was vital for predictive analytics and process optimization, while " Big data analytics for smart decision-making (S1)"—despite its importance—ranked fifth. Urban managers should prioritize the development of IOT networks to fully leverage their potential for resource management and efficiency gains. Following this, attention to process automation and AI integration can significantly enhance the quality of life for citizens and reduce urban costs.

Model‐Driven Manufacturing of High‐Energy‐Density Batteries: A Review

AbstractThe rapid advancement in energy storage technologies, particularly high‐energy density batteries, is pivotal for diverse applications ranging from portable electronics to electric vehicles and grid storage. This review paper provides a comprehensive analysis of the recent progress in model‐driven manufacturing approaches for high‐energy‐density batteries, highlighting the integration of computational models and simulations with experimental manufacturing processes to optimize performance, reliability, safety, and cost‐effectiveness. We systematically examine various modeling techniques, including electrochemical, thermal, and mechanical models, and their roles in elucidating the complex interplay of materials, design, and manufacturing parameters. The review also discusses the challenges and opportunities in scaling up these model‐driven approaches, addressing key issues such as model validation, parameter sensitivity, and the integration of machine learning and artificial intelligence for predictive modeling, process optimization, and quality assurance. By synthesizing current research findings and industry practices, this paper aims to outline a roadmap for future developments in model‐driven manufacturing of high‐energy density batteries, emphasizing the need for interdisciplinary collaboration and innovation to meet the increasing demands for energy storage solutions.

High-level visual prediction errors in early visual cortex.

Perception is shaped by both incoming sensory input and expectations derived from our prior knowledge. Numerous studies have shown stronger neural activity for surprising inputs, suggestive of predictive processing. However, it is largely unclear what predictions are made across the cortical hierarchy, and therefore what kind of surprise drives this up-regulation of activity. Here, we leveraged fMRI in human volunteers and deep neural network (DNN) models to arbitrate between 2 hypotheses: prediction errors may signal a local mismatch between input and expectation at each level of the cortical hierarchy, or prediction errors may be computed at higher levels and the resulting surprise signal is broadcast to earlier areas in the cortical hierarchy. Our results align with the latter hypothesis. Prediction errors in both low- and high-level visual cortex responded to high-level, but not low-level, visual surprise. This scaling with high-level surprise in early visual cortex strongly diverged from feedforward tuning. Combined, our results suggest that high-level predictions constrain sensory processing in earlier areas, thereby aiding perceptual inference.

Predictive learning shapes the representational geometry of the human brain

Predictive coding theories propose that the brain constantly updates internal models to minimize prediction errors and optimize sensory processing. However, the neural mechanisms that link prediction error encoding and optimization of sensory representations remain unclear. Here, we provide evidence how predictive learning shapes the representational geometry of the human brain. We recorded magnetoencephalography (MEG) in humans listening to acoustic sequences with different levels of regularity. We found that the brain aligns its representational geometry to match the statistical structure of the sensory inputs, by clustering temporally contiguous and predictable stimuli. Crucially, the magnitude of this representational shift correlates with the synergistic encoding of prediction errors in a network of high-level and sensory areas. Our findings suggest that, in response to the statistical regularities of the environment, large-scale neural interactions engaged in predictive processing modulate the representational content of sensory areas to enhance sensory processing.

How the prospect of a clinical trial impacts decision-making for predictive genetic testing in ALS

Objective: Genetic testing practices are rapidly evolving for people living with, or at-risk for, amyotrophic lateral sclerosis (ALS), due to emerging genotype-driven therapies. This study explored how individuals at-risk for familial ALS (fALS) perceive the opportunity to participate in a clinical trial, and to better understand how that may influence the decision-making process for predictive genetic testing. Methods: This study used both quantitative and qualitative data analyses. Data were collected through an online questionnaire, followed by semi-structured interviews conducted with twelve (n = 12) individuals at-risk for either SOD1- or C9orf72-ALS who had predictive testing prior to study participation. Interview data were analyzed using reflexive thematic analysis. Results: Three overarching themes were conceptualized from the data: i) the psychosocial impact of fALS; ii) perspectives of at-risk individuals on research involvement; and iii) predictive genetic counseling and testing considerations. These results contribute perspectives of the lived experience to inform predictive genetic counseling and testing practices for individuals at-risk for fALS. Conclusion: Individuals at-risk for fALS view potential participation in a presymptomatic clinical trial as an actionable measure that may increase their desire for predictive genetic testing. Genetic counseling was identified as a critical component of the predictive testing process given the life-changing implications associated with a positive result. Increased access to predictive genetic counseling, and in a timely manner, is a significant need in the ALS population given potential access to gene-specific therapies in the presymptomatic stage.