Inference Model Research Articles

Generalized Geographically Weighted Regression Model within a Modularized Bayesian Framework.

Geographically weighted regression (GWR) models handle geographical dependence through a spatially varying coefficient model and have been widely used in applied science, but its general Bayesian extension is unclear because it involves a weighted log-likelihood which does not imply a probability distribution on data. We present a Bayesian GWR model and show that its essence is dealing with partial misspecification of the model. Current modularized Bayesian inference models accommodate partial misspecification from a single component of the model. We extend these models to handle partial misspecification in more than one component of the model, as required for our Bayesian GWR model. Information from the various spatial locations is manipulated via a geographically weighted kernel and the optimal manipulation is chosen according to a Kullback-Leibler (KL) divergence. We justify the model via an information risk minimization approach and show the consistency of the proposed estimator in terms of a geographically weighted KL divergence.

Development of the edible Tenebrio molitor at different temperatures: a Poisson log-linear regression modeling approach

AbstractTenebrio molitor gained recognition for its nutritional value for human and/or animal consumption, its utilization as a model species in research studies, and its ability to degrade many types of plastics. However, it is a stored-product pest infesting a wide range of commodities in storages. The impact of temperature on the development of T. molitor was evaluated, taking into account confounding effects of other covariates such as developmental stage and larval instar. The time for larval development was longer at 22.5 °C than at 25 °C, 27.5 °C, and 30 °C. Tenebrio molitor spent most of its lifetime as larva > egg, pupa, since the outcome of our model inference resulted that the estimated probability of longer duration in larvae is approximately 25.5 times higher compared to eggs and pupae, and passed through 26 instars before it became pupa. Duration of larval instars increased significantly up to L22 and then decreased until L26. Among larval instars, L22 exhibited the highest developmental duration while the lowest was recorded for L1. These findings enhance the knowledge about the biology of T. molitor, leading to precise decisions for its successful culture.

Tricks with metrics: combining statistics for improved inference in regression analysis

. In maximum likelihood methods, the three classical tests statistics are often unreliable for inference in small samples, even under the correct model specification. In this article, I discuss how likelihood ratio and Wald test statistics can be combined to obtain highly improved parameter inference in regression models with small samples. I consider modifications obtained from both the Barndorff-Nielsen and the Lugannani and Rice likelihood approximations, and I show how they can produce highly accurate parameter inference in a general (possibly non linear) regression model with possibly non spherical disturbances. I discuss the underlying theory and provide Monte Carlo simulations demonstrating the superior accuracy of the proposed procedures over the first-order classical likelihood methods (i.e., the signed log-likelihood ratio test and the Wald test). An empirical application to a regression model of mobile money (“M-pesa”) adoption in Kenya is provided as an illustration of the usefulness of these methods in practice.

Fast Variational Inference for Bayesian Factor Analysis in Single and Multi-Study Settings

– Factors models are commonly used to analyze high-dimensional data in both single-study and multi-study settings. Bayesian inference for such models relies on Markov Chain Monte Carlo (MCMC) methods, which scale poorly as the number of studies, observations, or measured variables increase. To address this issue, we propose new variational inference algorithms to approximate the posterior distribution of Bayesian latent factor models using the multiplicative gamma process shrinkage prior. The proposed algorithms provide fast approximate inference at a fraction of the time and memory of MCMC-based implementations while maintaining comparable accuracy in characterizing the data covariance matrix. We conduct extensive simulations to evaluate our proposed algorithms and show their utility in estimating the model for high-dimensional multi-study gene expression data in ovarian cancers. Overall, our proposed approaches enable more efficient and scalable inference for factor models, facilitating their use in high-dimensional settings. An R package VIMSFA implementing our methods is available on GitHub (github.com/blhansen/VI-MSFA).

Making the most of AI and machine learning in organizations and strategy research: Supervised machine learning, causal inference, and matching models

AbstractResearch SummaryWe spotlight the use of machine learning in two‐stage matching models to deal with sample selection bias. Recent advances in machine learning have unlocked new empirical possibilities for inductive theorizing. In contrast, the opportunities to use machine learning in regression studies involving large‐scale data with many covariates and a causal claim are still less well understood. Our core contribution is to guide researchers in the use of machine learning approaches to choosing matching variables for enhanced causal inference in propensity score matching models. We use an analysis of real‐world technology invention data of public–private relationships to demonstrate the method and find that machine learning can provide an alternative approach to ad hoc matching. However, as with any method, it is also important to understand its limitations.Managerial SummaryThis article explores the use of machine learning to enhance decision‐making, particularly in addressing sample selection bias in large‐scale datasets. The rapid development of AI and machine learning offers new, powerful tools especially for digital ecosystems where complex data and causal relationships are complex to analyze. We offer managers and stakeholders insight into the effective integration of machine learning for selecting critical variables in propensity score matching models. Through a detailed examination of real‐world data on technology inventions within public–private relationships, we demonstrate the effectiveness of machine learning as a robust alternative to traditional matching methods.

Connecting process models to response times through Bayesian hierarchical regression analysis.

Process models specify a series of mental operations necessary to complete a task. We demonstrate how to use process models to analyze response-time data and obtain parameter estimates that have a clear psychological interpretation. A prerequisite for our analysis is a process model that generates a count of elementary information processing steps (EIP steps) for each trial of an experiment. We can estimate the duration of an EIP step by assuming that every EIP step is of random duration, modeled as draws from a gamma distribution. A natural effect of summing several random EIP steps is that the expected spread of the overall response time increases with a higher EIP step count. With modern probabilistic programming tools, it becomes relatively easy to fit Bayesian hierarchical models to data and thus estimate the duration of a step for each individual participant. We present two examples in this paper: The first example is children's performance on simple addition tasks, where the response time is often well predicted by the smaller of the two addends. The second example is response times in a Sudoku task. Here, the process model contains some random decisions and the EIP step count thus becomes latent. We show how our EIP regression model can be extended to such a case. We believe this approach can be used to bridge the gap between classical cognitive modeling and statistical inference and will be easily applicable to many use cases.

In silico model development and optimization of in vitro lung cell population growth.

Tissue engineering predominantly relies on trial and error in vitro and ex vivo experiments to develop protocols and bioreactors to generate functional tissues. As an alternative, in silico methods have the potential to significantly reduce the timelines and costs of experimental programs for tissue engineering. In this paper, we propose a methodology to formulate, select, calibrate, and test mathematical models to predict cell population growth as a function of the biochemical environment and to design optimal experimental protocols for model inference of in silico model parameters. We systematically combine methods from the experimental design, mathematical statistics, and optimization literature to develop unique and explainable mathematical models for cell population dynamics. The proposed methodology is applied to the development of this first published model for a population of the airway-relevant bronchio-alveolar epithelial (BEAS-2B) cell line as a function of the concentration of metabolic-related biochemical substrates. The resulting model is a system of ordinary differential equations that predict the temporal dynamics of BEAS-2B cell populations as a function of the initial seeded cell population and the glucose, oxygen, and lactate concentrations in the growth media, using seven parameters rigorously inferred from optimally designed in vitro experiments.

Severe outcomes of malaria in children under time-varying exposure

In malaria epidemiology, interpolation frameworks based on available observations are critical for policy decisions and interpreting disease burden. Updating our understanding of the empirical evidence across different populations, settings, and timeframes is crucial to improving inference for supporting public health. Here, via individual-based modeling, we evaluate a large, multicountry, contemporary Plasmodium falciparum severe malaria dataset to better understand the relationship between prevalence and incidence of malaria pediatric hospitalizations - a proxy of malaria severe outcomes- in East-Africa. We find that life-long exposure dynamics, and subsequent protection patterns in children, substantially determine the likelihood of malaria hospitalizations relative to ongoing prevalence at the population level. Unsteady transmission patterns over a lifetime in children -increasing or decreasing- lead to an exponential relationship of hospitalization rates versus prevalence rather than the asymptotic pattern observed under steady transmission. Addressing this increase in the complexity of malaria epidemiology is crucial to update burden assessments via inference models that guide current and future policy decisions.

On the generalization of cognitive optical networking applications using composable machine learning

Model generalization characterizes the sustainability of machine learning (ML) designs applied to novel system states and therefore plays a vital role toward the realization of cognitive networking. In this paper, we present a composable ML framework (namely, CompML), aiming at generalizing ML-aided cognitive applications for optical networks. CompML makes use of three basic functional modules, i.e., the Loading, Recursion, and Readout modules, to model the loading/initialization processes (e.g., the launch of a signal), extract cumulative features by recursive operations, and produce model inferences, respectively. By the composition of the three modules and adoption of an end-to-end training mechanism, CompML allows for generalizing multiple tasks of the same domain [e.g., quality-of-transmission (QoT) estimation for different lightpaths]. We perform case studies of CompML on QoT estimation and nonlinearity compensation using both simulation and experimental data. Results show the superior generalization ability of CompML compared with the baselines, achieving mean absolute error (MAE) for generalized signal-to-noise ratio (GSNR) prediction error of below 1.06 dB for unseen lightpaths and up to 3 dB Q-factor improvement for nonlinearity compensation.

Federated variational generative learning for heterogeneous data in distributed environments

Distributedly training models across diverse clients with heterogeneous data samples can significantly impact the convergence of federated learning. Various novel federated learning methods address these challenges but often require significant communication resources and local computational capacity, leading to reduced global inference accuracy in scenarios with imbalanced label data distribution and quantity skew. To tackle these challenges, we propose FedVGL, a Federated Variational Generative Learning method that directly trains a local generative model to learn the distribution of local features and improve global target model inference accuracy during aggregation, particularly under conditions of severe data heterogeneity. FedVGL facilitates distributed learning by sharing generators and latent vectors with the global server, aiding in global target model training from mapping local data distribution to the variational latent space for feature reconstruction. Additionally, FedVGL implements anonymization and encryption techniques to bolster privacy during generative model transmission and aggregation. In comparison to vanilla federated learning, FedVGL minimizes communication overhead, demonstrating superior accuracy even with minimal communication rounds. It effectively mitigates model drift in scenarios with heterogeneous data, delivering improved target model training outcomes. Empirical results establish FedVGL's superiority over baseline federated learning methods under severe label imbalance and data skew condition. In a Label-based Dirichlet Distribution setting with α=0.01 and 10 clients using the MNIST dataset, FedVGL achieved an exceptional accuracy over 97% with the VGG-9 target model.

Trending Time-Varying Coefficient Spatial Panel Data Models

This article investigates the estimation and inference of spatial panel data models in which the regression coefficient vector is a trending function. We use time differences to eliminate the individual effects and employ various GMM estimations for regression coefficients with both linear and quadratic moments. Time trend estimator based on these GMM estimations is also proposed. Monte Carlo experiments show that the finite sample performance of the estimators is satisfactory. As an empirical illustration, we investigate the trending pattern of the spillover effect of air pollution among Chinese cities from 2015 to 2021.

Fine-grained recognition of bitter gourd maturity based on Improved YOLOv5-seg model

Bitter gourd, being perishable, requires timely harvesting. Delayed harvesting can result in a substantial reduction in fruit quality. while premature harvesting leads to underdeveloped fruit and decreased yields, the continuous flowering pattern in bitter gourd underscores the significance of accurately assessing fruit growth and ensuring timely harvesting for subsequent fruit setting and development. The current reliance on the experience of production personnel represents a substantial inefficiency. We present an improved real-time instance segmentation model based on YOLOv5-seg. The utilization of dynamic snake convolution enables the extraction of morphological features from the curved and elongated structure of bitter gourd. Diverse branch blocks enhance feature space diversity without inflating model size and inference time, contributing to improved recognition of expansion stages during bitter gourd growth. Additionally, the introduction of Focal-EIOU loss accurately locates the boundary box and mask, addressing sample imbalances in the L2 stage. Experimental results showcase remarkable accuracy rates of 99.3%, 93.8%, and 98.3% for L1, L2, and L3 stages using mAP@0.5. In comparison, our model outperforms other case segmentation models, excelling in both detection accuracy and inference speed. The improved YOLOv5-seg model demonstrates strong performance in fine-grained recognition of bitter gourd during the expansion stage. It efficiently segments bitter gourd in real-time under varying lighting and occlusion conditions, providing crucial maturity information. This model offers reliable insights for agricultural workers, facilitating precise harvesting decisions.

Evaluating the predictions of an interoceptive inference model of bulimia nervosa

ObjectiveBulimia nervosa (BN) is associated with loss-of-control (LOC) eating episodes that frequently occur in response to negative emotions. According to recent neurocomputational models, this link could be explained by a failure to accurately update beliefs about the body in states of high arousal. Specifically, these interoceptive inference models suggest that under-relying on signals from one’s body about sensory experience (“low sensory precision”) and/or over-relying on previously held beliefs (“excessively precise priors”) lead to inaccurate perception and maladaptive behaviors. We conducted an initial test of these core predictions of the interoceptive inference model in BN using self-report measures.MethodsWe compared women with BN (n = 30) and age-, BMI-, and full-scale IQ-matched controls (n = 31) on trust in sensory information from the body and two types of beliefs about what can be done to regulate high negative affect. Within the BN group, we tested interrelations among these measures and explored their associations with LOC eating frequency.ResultsCompared with healthy controls, the BN group reported lower levels of trust in sensory information and stronger beliefs that once upset, there is little one can do, apart from eating, to self-regulate. These beliefs were associated with each other and with lower body trust. Beliefs about the uncontrollability of emotion were associated with more frequent subjective binge-eating episodes.ConclusionsFindings provide initial support for the core predictions of an interoceptive inference account of BN: low trust in sensory information (“sensory precision”) may promote an overreliance on maladaptive “prior beliefs” about the effects of eating on negative emotions, ultimately interfering with accurate updating of beliefs about other strategies that could regulate emotions and maintain LOC eating. Low body trust, strong expectations about emotions, and their neurocomputational underpinnings could be promising combined treatment targets for BN.

A deep neural network model for Chinese toponym matching with geographic pre-training model

ABSTRACT Multiple tasks within the field of geographical information retrieval and geographical information sciences necessitate toponym matching, which involves the challenge of aligning toponyms that share a common referent. The multiple string similarity approaches struggle when confronted with the complexities associated with unofficial and/or historical variants of identical toponyms. Also, current state-of-the-art approaches/tools to supervised machine learning rely on labeled samples, and they do not adequately address the intricacies of character replacements either from transliterations or historical shifts in linguistic and cultural norms. To address these issues, this paper proposes a novel matching approach that leverages a deep neural network model empowered by geographic language representation model, known as GeoBERT, which stands for geographic Bidirectional Encoder Representations from Transformers (BERT). This model harnesses the groundbreaking capabilities of the GeoBERT framework by extending a generalized Enhanced Sequential Inference Model architecture and integrating multiple features to enhance the accuracy and robustness of the toponym matching. We present a comprehensive evaluation of the proposed method’s performance using three extensive datasets. The findings clearly illustrate that our approach outperforms the individual similarity metrics used in previous studies.

A Smart Application Frame of Remote Sensing in Non-grain Production Data Governance

Abstract. This study addresses the intricate challenges encountered in the data governance process of Non-grain Production (NGP) on Arable land. This involves managing data from diverse sources, with varying accuracies and formats, and utilizing multiple specialized software tools. An object-oriented approach is adopted to encapsulate experiential knowledge related to the data and associated processing methods, thus creating an Application Knowledge Body Model (AKBM). This model acts as a conduit between users and computational resources, encompassing various types of data and their corresponding processing and analysis methods. Moreover, by employing model inference techniques to devise methods for transitioning from raw data models to target models, a foundation is laid for the accumulation, sharing, and intelligent application of expertise on data, methods, models, and knowledge.The application examples demonstrate that users can directly construct new solutions containing relevant data and associated processing methods, rather than grappling with a multitude of data files and complex specialized software when encountering novel challenges. This promotes collaborative development in data governance on geospatial big data platforms, significantly enhancing governance efficiency, improving the quality of information support in NGP cultivation management, advancing current technological capabilities, and fostering the progression of related technologies.

Distribution-free Inferential Models: Achieving finite-sample valid probabilistic inference, with emphasis on quantile regression

This paper presents a novel distribution-free Inferential Model (IM) construction that provides valid probabilistic inference across a broad spectrum of distribution-free problems, even in finite sample settings. More specifically, the proposed IM has the capability to assign (imprecise) probabilities to assertions of interest about any feature of the unknown quantities under examination, and these probabilities are well-calibrated in a frequentist sense. It is also shown that finite-sample confidence regions can be derived from the IM for any such features. Particular emphasis is placed on quantile regression, a domain where uncertainty quantification often takes the form of set estimates for the regression coefficients in applications. Within this context, the IM facilitates the acquisition of these set estimates, ensuring they are finite-sample confidence regions. It also enables the provision of finite-sample valid probabilistic assignments for any assertions of interest about the regression coefficients. As a result, regardless of the type of uncertainty quantification desired, the proposed framework offers an appealing solution to quantile regression.

Logistics integration between the Manaus Free Trade Zone (ZFM) and Venezuela: evaluation of the modal through the Fuzzy logic for the flow of production in the electronics sector

The Logistics Integration between the Manaus Free Trade Zone (ZFM) and Venezuela regarding the flow of production in the electrical and electronic sector, has specificities in its commercial activity. Competition and the rapid increase in consumption form a more competitive environment, where the ability to meet fluctuating demand in time stands as a central component of competitive advantage. The main objective of this research was: To develop a Fuzzy inference model for the evaluation of the most viable modal for the integration between the Manaus Free Trade Zone (ZFM) and Venezuela in the flow of production in the electronics sector. The Methodological Process was developed in three phases: 1. Identification of the indicators for the Evaluation of Available Modals; 2. Fuzzy "Inference" System Modeling; 3. Experiment of the Proposed Model, with its respective stages: Phase 1 (definition of the inference levels of the indicators), Phase 2 (Development of Fuzzy Sets, Development of the "Inference" and Simulation Rules in the MatlabR2013 software), Phase 3 (Compilation of the Indicator Aggregation Algorithm, Simulation of the Results in 3D and Analysis of the Results Obtained). Based on the observed data from the 3D results, in figure 16, a detailed analysis of the behavior of the variables was allowed. The results obtained with the proposed model were satisfactory in the analysis of the performance classifications of the modes, it was possible to identify the most relevant variables for the flow of the production of the electrical and electronic sector between the Manaus Free Trade Zone and the Venezuela.

Generalized data thinning using sufficient statistics

Our goal is to develop a general strategy to decompose a random variable X into multiple independent random variables, without sacrificing any information about unknown parameters. A recent paper showed that for some well-known natural exponential families, X can be thinned into independent random variables X ( 1 ) , … , X ( K ) , such that X = ∑ k = 1 K X ( k ) . These independent random variables can then be used for various model validation and inference tasks, including in contexts where traditional sample splitting fails. In this paper, we generalize their procedure by relaxing this summation requirement and simply asking that some known function of the independent random variables exactly reconstruct X. This generalization of the procedure serves two purposes. First, it greatly expands the families of distributions for which thinning can be performed. Second, it unifies sample splitting and data thinning, which on the surface seem to be very different, as applications of the same principle. This shared principle is sufficiency. We use this insight to perform generalized thinning operations for a diverse set of families.

Modeling- and Simulation-Driven Methodology for the Deployment of an Inland Water Monitoring System

In response to the challenges introduced by global warming and increased eutrophication, this paper presents an innovative modeling and simulation (M&S)-driven model for developing an automated inland water monitoring system. This system is grounded in a layered Internet of Things (IoT) architecture and seamlessly integrates cloud, fog, and edge computing to enable sophisticated, real-time environmental surveillance and prediction of harmful algal and cyanobacterial blooms (HACBs). Utilizing autonomous boats as mobile data collection units within the edge layer, the system efficiently tracks algae and cyanobacteria proliferation and relays critical data upward through the architecture. These data feed into advanced inference models within the cloud layer, which inform predictive algorithms in the fog layer, orchestrating subsequent data-gathering missions. This paper also details a complete development environment that facilitates the system lifecycle from concept to deployment. The modular design is powered by Discrete Event System Specification (DEVS) and offers unparalleled adaptability, allowing developers to simulate, validate, and deploy modules incrementally and cutting across traditional developmental phases.

Improved YOLOv8-Pose Algorithm for Albacore Tuna (Thunnus alalunga) Fork Length Extraction and Weight Estimation

Aiming at the problems of large statistical error and the poor real-time performance of catch weight in the ocean fishing tuna industry, an algorithm based on improved YOLOv8-Pose for albacore tuna (Thunnus alalunga) fork length extraction and weight estimation is proposed, with reference to the human body’s pose estimation algorithm. Firstly, a lightweight module constructed using a heavy parameterization technique is used to replace the backbone network, and secondly, a weighted bidirectional feature pyramid network BIFPN is utilized. Finally, the upper and lower jaw and tail feature points of the albacore tuna (Thunnus alalunga) were extracted using the key point detection algorithm, and the weight of the albacore tuna (Thunnus alalunga) was estimated based on the fitted relationship between fork length and weight. The experimental results show that the improved YOLOv8-Pose algorithm reduces the number of model parameters by 13.63% and the number of floating-point operations by 14.03% compared with the baseline model without decreasing the accuracy of the target detection and key point detection and improves the model inference speed by 374%. At the same time, it reduces the drift of the key point detection, and the error of the comparison with the actual albacore tuna (Thunnus alalunga) body weight is not more than 10%. The improved key point detection algorithm has high detection accuracy and inference speed, which provides accurate yield data for pelagic fishing and is expected to solve the existing statistical problems and improve the accuracy and real-time performance of data in the fishing industry.