Estimation Performance Research Articles

Inclusion of genotype by environment interactions into genomic prediction models for sorghum seed parents in hybrid testcrosses

ABSTRACT Multi-environment trials are routinely used to screen sorghum (Sorghum bicolor) hybrids for quantitative traits, such as grain yield (GY) and days to mid-anthesis (DY). By nature, quantitative traits are greatly affected by environmental effects, and it is common that hybrids have unequal relative performance across environments. Referred to as genotype by environment (G×E) interaction effects, unequal relative performance hinders the selection of superior hybrids. This study assesses the efficacy of including G×E effects into genomic prediction models of sorghum hybrid performance for GY and DY across environments. Testcross hybrids from two B-line populations were evaluated across multiple environments to generate predictions in four different cross validation (CV) schemes. These were CV1, individual environment predictions; CV2, multi-environment predictions where the predicted hybrids were not observed in any environment; CV3, multi-environment predictions where the predicted hybrids were sparsely tested across environments; and CV4, where a grand best linear unbiased estimate (BLUE) of hybrid performance across environments was predicted. The inclusion of G×E effects into genomic prediction models was no more predictive under a CV2 scheme than individual environment genomic prediction models (CV1). However, genomic prediction models that included G×E effects in a CV3 scheme improved prediction accuracies for both GY and DY. The overall predictive ability of CV4 was also higher than that of CV1 and CV2, but there was no clear difference between CV4 and CV3. The results herein provide a framework on how genomic prediction can be incorporated in sorghum breeding programs to predict quantitative traits and increase genetic gain.

C 3 -GAN+: Complex-Condition-Controlled Generative Adversarial Networks with Enhanced Embedding

Given historical traffic distributions and associated urban conditions observed in a city, the conditional urban traffic estimation problem aims at estimating realistic future projections of the traffic under a set of new urban conditions, e.g., new bus routes, rainfall intensity, and travel demands. The problem is important in reducing traffic congestion, improving public transportation efficiency, and facilitating urban planning. However, solving this problem is challenging due to the strong spatial dependencies of traffic patterns and the complex relations between the traffic and urban conditions. Recently, we proposed a Complex-Condition-Controlled Generative Adversarial Network ( \(\boldsymbol{C^{3}}\) -GAN) , which tackles both of the challenges and solves the urban traffic estimation problem under various complex conditions by adding a fixed embedding network and an inference network on top of the standard conditional GAN model. The randomly chosen embedding network transforms the complex conditions to latent vectors, and the inference network enhances the connections between the embedded vectors and the traffic data. However, a randomly chosen embedding network cannot always successfully extract features of complex urban conditions, which indicates \(C^{3}\) -GAN is unable to uniquely map different urban conditions to proper latent distributions. Thus, \(C^{3}\) -GAN would fail in certain traffic estimation tasks. Besides, \(C^{3}\) -GAN is hard to train due to vanishing gradients and mode collapse problems. To address these issues, in this article, we extend our prior work by introducing a new deep generative model, namely, \(C^{3}\) -GAN \(+\) , which significantly improves the estimation performance and model stability. \(C^{3}\) -GAN \(+\) has new objective, architecture, and training algorithm. The new objective applies Wasserstein loss to the conditional generation case to encourage stable training. Shared convolutional layers between the discriminator and the inference network help to capture spatial dependencies of traffic more efficiently, part of the shared convolutional layers are used to update the embedding network periodically aiming to encourage good representation and avoid model divergence. Extensive experiments on real-world datasets demonstrate that our \(C^{3}\) -GAN \(+\) produces high-quality traffic estimations and outperforms state-of-the-art baseline methods.

Assessing the impact of pedigree attributes on the validity of quantitative genetic parameter estimates.

Investigating the evolution of complex traits in nature requires accurate assessment of their genetic basis. Quantitative genetic (QG) modeling is frequently applied to estimate the additive genetic variance (VA) in traits, combining phenotypic and pedigree data from a sample of individuals. Whether reconstructed from social links or molecular markers, empirical pedigrees differ in completeness, genealogical error rates and other attributes that can impact QG estimation. Here we investigate this impact using human genealogical data for six French-Canadian (FC) populations originating from the same genetic founding source but differing in their pedigrees' attributes. First, we simulated phenotypic values along pedigrees and under different trait architecture and 'true' parameter values (e.g. VA). Then we fitted mixed effects 'animal' models to these simulated data, to assess how QG estimation was impacted by pedigree attributes. Our results show that pedigree size and depth were important determinants of the precision, but not accuracy, of genetic parameter estimates. In contrast, pedigree completeness and entropy, two attributes related to the density of genealogical links, were not clearly associated with the performance of parameter estimation. Noticeably, a slight increase in the genealogical error rate was sufficient to cause a detectable underestimation of VA. Including maternal genetic effects into the simulations lead to a slight underestimation of VA with pedigrees of smaller size and depth. Despite originating from the same genetic source, the six pedigrees yielded wide variations in QG estimates under identical conditions. These findings highlight the importance of sensitivity analyses in pedigree-based genetic studies on natural populations.

Using Night-Time Drone-Acquired Thermal Imagery to Monitor Flying-Fox Productivity—A Proof of Concept

Accurate and precise monitoring of species abundance is essential for determining population trends and responses to environmental change. Species, such as bats, that have slow life histories, characterized by extended lifespans and low reproductive rates, are particularly vulnerable to environmental changes, stochastic events, and human activities. An accurate assessment of productivity can improve parameters for population modelling and provide insights into species’ capacity to recover from population perturbations, yet data on reproductive output are often lacking. Recently, advances in drone technology have allowed for the development of a drone-based thermal remote sensing technique to accurately and precisely count the numbers of flying-foxes (Pteropus spp.) in their tree roosts. Here, we extend that method and use a drone-borne thermal camera flown at night to count the number of flying-fox pups that are left alone in the roost whilst their mothers are out foraging. We show that this is an effective method of estimating flying-fox productivity on a per-colony basis, in a standardized fashion, and at a relatively low cost. When combined with a day-time drone flight used to estimate the number of adults in a colony, this can also provide an estimate of female reproductive performance, which is important for assessments of population health. These estimates can be related to changes in local food availability and weather conditions (including extreme heat events) and enable us to determine, for the first time, the impacts of disturbances from site-specific management actions on flying-fox population trajectories.

Estimation and Bayesian Prediction for the Generalized Exponential Distribution Under Type-II Censoring

This research focuses on the prediction and estimation problems for the generalized exponential distribution under Type-II censoring. Firstly, maximum likelihood estimations for the parameters of the generalized exponential distribution are computed using the EM algorithm. Additionally, confidence intervals derived from the Fisher information matrix are developed and analyzed alongside two bootstrap confidence intervals for comparison. Compared to classical maximum likelihood estimation, Bayesian inference proves to be highly effective in handling censored data. This study explores Bayesian inference for estimating the unknown parameters, considering both symmetrical and asymmetrical loss functions. Utilizing Gibbs sampling to produce Markov Chain Monte Carlo samples, we employ an importance sampling approach to obtain Bayesian estimates and compute the corresponding highest posterior density (HPD) intervals. Furthermore, for one-sample prediction and, separately, for the two-sample case, we provide the corresponding posterior distributions, along with methods for computing point predictions and predictive intervals. Through Monte Carlo simulations, we evaluate the performance of Bayesian estimation in contrast to maximum likelihood estimation. Finally, we conduct an analysis of a real dataset derived from deep groove ball bearings, calculating Bayesian point predictions and predictive intervals for future samples.

Stochastic encryption against stealthy attacks in CPSs: A zero-sum game approach

Stochastic encryption against stealthy attacks in CPSs: A zero-sum game approach

Nonparametric Threshold Estimation of Autocorrelated Statistics in Multivariate Statistical Process Monitoring

ABSTRACTMultivariate statistical process monitoring is commonly used to detect abnormal process behavior in real time. Multiple process variables are monitored simultaneously, and alarms are issued when monitoring statistics exceed a predetermined threshold. Traditional approaches use a parametric threshold based on the assumptions of independence and multivariate normality of the process data, which are often violated in complex processes with high sampling frequencies, leading to excessive false alarms. Some approaches for improved threshold selection have been proposed, but they assume independence of the monitoring statistics, which are often autocorrelated. In this paper, we compare the performance of nonparametric estimators for computing thresholds from autocorrelated monitoring statistics through simulation. The false alarm rate and in‐control average run length of each estimator under different distributions, sample sizes, and autocorrelation levels and types are found. Estimator performance is found to depend on sample size and the strength of autocorrelation. The class of kernel density estimation (KDE) methods tends to perform better than estimators that use bootstrapping, and the proposed adjusted KDE methods that account for autocorrelation are recommended for general use. A case study to monitor a wastewater treatment facility further illustrates the performance of nonparametric and parametric thresholds when applied to real‐world systems.

GTIGNet: Global Topology Interaction Graphormer Network for 3D hand pose estimation.

Estimating 3D hand poses from monocular RGB images presents a series of challenges, including complex hand structures, self-occlusions, and depth ambiguities. Existing methods often fall short of capturing the long-distance dependencies of skeletal and non-skeletal connections for hand joints. To address these limitations, we introduce the Global Topology Interaction Graphormer Network (GTIGNet), a novel deep learning architecture designed to improve 3D hand pose estimation. Our model incorporates a Context-Aware Attention Block (CAAB) within the 2D pose estimator to enhance the extraction of multi-scale features, yielding more accurate 2D joint heatmaps to support the task that followed. Additionally, we introduce a High-Order Graphormer that explicitly and implicitly models the topological structure of hand joints, thereby enhancing feature interaction. Ablation studies confirm the effectiveness of our approach, and experimental results on four challenging datasets, Rendered Hand Dataset (RHD), Stereo Hand Pose Benchmark (STB), First-Person Hand Action Benchmark (FPHA), and FreiHAND Dataset, indicate that GTIGNet achieves state-of-the-art performance in 3D hand pose estimation. Notably, our model achieves an impressive Mean Per Joint Position Error (MPJPE) of 9.98 mm on RHD, 6.12 mm on STB, 11.15 mm on FPHA and 10.97 mm on FreiHAND.

Intelligent Strategic Planning Method based Algorithm (ISPMA) for Estimation of Soccer Sports Match Outcome

Estimating the soccer match outcome with adequate accuracy is still one of the biggest challenges in the sports domain. In this work, the proposed novel Intelligent Strategic Planning Method based Algorithm (ISPMA) carries out dynamic estimation of soccer team performance in terms of the match outcome and noticeably outperforms the existing state of the art methods due to its unique features. In this work, the output of the four feature selection machine learning techniques i.e. Pearson correlation, forward selection, Extra tree classifier, and CHI-square is firstly unified before feeding these selected features as an input to the seven classifiers i.e. SVM (Support Vector Machine), Naïve Bayes, KNN (K-Nearest Neighbor.), Decision Tree, Random Forest, Logistic Regression, and AdaBoost. The dataset comprises eleven seasons of the English premier league and 3762 matches have been used to train the model and 418 matches to test the same. Such a reasonable size of soccer dataset is not common in previous studies. Another unique feature of this work is the time of estimation as estimation can be done during the progression of the game based on match statistics associated with the first half of the match. The proposed method uses a novel approach by computing the average values of the selected set of features for the victory of the team to estimate match results. By using these computed average values, ISPMA generates strategic planning based suggestions for the second half of the match. The strategic planning generated by the proposed method facilitates estimating the team performance and shifting the momentum from one team to another and can assist the coach, managers, and the team in carrying out effective decision-making for better match outcome.

Estimates of Wind Speed Profiles from Surface Observations: Machine Learning Versus Monin–Obukhov Approach

Accurate estimation of near-surface wind profiles with surface observations is important for evaluating wind resources. Based on the multilayer perceptron algorithm, this study proposes a machine learning (ML) model by establishing the relationship between instantaneous near-surface atmospheric stability (e.g., exponent α in the power law method) and the mean, change and standard deviation values of the wind speed at 10 m, temperature and relative humidity at 2 m in the past hours. The evaluation results for six wind tower sites in semiarid and arid regions of northwestern China indicate that, compared with the Monin–Obukhov method (e.g., Holtslag in Boundary-Layer Meteorol 29:225–250, 1984), the mean relative errors of near-surface wind speed in the ML model could be reduced by 8.4%, 10.6% and 8.3% at 30 m, 50 m and 70 m, respectively. Further investigations suggest that compared to the MO model, the mean relative errors of near-surface wind profiles in the ML model could be reduced by 2–15% at different wind tower sites and under different near-surface stability conditions. In general, the estimation performance of near-surface wind profiles in the ML model is better under unstable conditions than under stable conditions. Results illuminate that the proposed method using near-surface variables in the past hours as inputs could be an effective way to improve the estimation of near-surface wind profiles.

6-DoF Pose Estimation from Single RGB Image and CAD Model Retrieval Using Feature Similarity Measurement

This study presents six degrees of freedom (6-DoF) pose estimation of an object from a single RGB image and retrieval of the matching CAD model by measuring the similarity between RGB and CAD rendering images. The 6-DoF pose estimation of an RGB object is one of the important techniques in 3D computer vision. However, in addition to 6-DoF pose estimation, retrieval and alignment of the matching CAD model with the RGB object should be performed for various industrial applications such as eXtended Reality (XR), Augmented Reality (AR), robot’s pick and place, and so on. This paper addresses 6-DoF pose estimation and CAD model retrieval problems simultaneously and quantitatively analyzes how much the 6-DoF pose estimation affects the CAD model retrieval performance. This study consists of two main steps. The first step is 6-DoF pose estimation based on the PoseContrast network. We enhance the structure of PoseConstrast by adding variance uncertainty weight and feature attention modules. The second step is the retrieval of the matching CAD model by an image similarity measurement between the CAD rendering and the RGB object. In our experiments, we used 2000 RGB images collected from Google and Bing search engines and 100 CAD models from ShapeNetCore. The Pascal3D+ dataset is used to train the pose estimation network and DELF features are used for the similarity measurement. Comprehensive ablation studies about the proposed network show the quantitative performance analysis with respect to the baseline model. Experimental results show that the pose estimation performance has a positive correlation with the CAD retrieval performance.

Performance of a blood-based test for colorectal cancer screening adjusted to the US census age and sex distribution.

18 Background: Despite availability of various screening options, nearly 40% of eligible U.S. adults were not up to date with the colorectal cancer (CRC) screening. Blood-based testing offers a promising complementary approach and may enhance patient adherence among unscreened individuals. In the PREEMPT CRC study, we evaluated clinical performance of an investigational blood-based screening test for detecting molecular signals of advanced colorectal neoplasia (ACN) in an average-risk population. Freenome blood-based CRC screening test met all primary endpoints demonstrating 79.2% sensitivity for CRC, 91.5% specificity for ACN, 90.8% negative predictive value (NPV) for ACN and 15.5% positive predictive value (PPV) for ACN. Sensitivity for advanced precancerous lesions (APLs) was 12.5%. To evaluate how the test would perform in the standard U.S. population, an analysis of test performance weighted to the U.S. census sex and age distribution was prespecified and performed. Methods: Between May 2020 and April 2022, the study enrolled participants aged 45-85 at average risk for CRC. Enrolled participants had a blood drawn before the bowel preparation for the standard of care screening colonoscopy (CS). CS and applicable histopathology reports underwent central pathologist review. Blood samples were processed blind to clinical findings. Primary endpoints included sensitivity for CRC, specificity for ACN, NPV for ACN and PPV for ACN. A secondary endpoint was sensitivity for APLs. Results: A subset of 32,731 sequentially enrolled participants were included in the clinical validation cohort. Out of those, 27,010 (82.5%) had evaluable blood samples and CS. The median age of the evaluable participants was 57.0 years, and 55.8% were women versus 50.5% in the U.S. Census. Age distribution included 11.0% participants in the 45-49 age group versus 15.3% in the U.S. Census, 33.0% in the 50-54 age group versus 15.6%, and 32.3% in the 55-64 age group versus 32.3%, 20.8% in the 65-74 age group versus 24.3%, 3.0% participants aged 75 or older versus 12.5%. After performance was weighted to match U.S. Census sex and age distributions, sensitivity for CRC was 81.1% (95% CI 71.3%-88.1%), specificity for those without ACN was 90.4% (95% CI 90.0%-90.7%), NPV for those without ACN was 90.5% (95% CI 90.5%-90.7%), and PPV for ACN was 15.5% (95% CI 13.8%-16.2%). Sensitivity for APL was 13.7% (95% CI 12.4%-15.0%). Conclusions: PREEMPT CRC is the largest study evaluating a CRC screening blood-based test to date. The study successfully met all primary endpoints, demonstrating acceptable clinical performance of the investigational blood-based test. The U.S. census sex-age adjusted clinical performance estimates showed improvement in CRC and APL sensitivity. This blood-based test may provide a convenient and effective option for CRC screening in the average-risk U.S. population. Clinical trial information: NCT04369053 .

Evaluation of the Impacts of Different Ground Motion Selection and Scaling Approaches on Seismic Performance of Bridges

ABSTRACT This paper investigates two ground motion selection and scaling (GMSS) methods for assessing the seismic performance of bridges in high seismic regions in California. It compares the traditional GMSS approach using elastic probabilistic seismic hazard analysis (PSHA) with a novel method based on inelastic PSHA, scaling ground motions to spectra across varying ductility levels. Nonlinear response history analyses of a 3-span bridge using 25 ground motions show that the inelastic PSHA-based method can reduce dispersion in responses, increasing confidence in seismic performance estimates. The traditional approach may produce conservative or unconservative results due to mismatches between scaled and target spectra.

Quantization tolerant network design and performance estimation of computation-in-memory for energy-efficient 3D object detection inference

Abstract This work discusses network architecture, network training method, and a quantization method for achieving a low-energy and high-energy-efficient system, aiming at equipping robots and automobiles such as Autonomous Mobile Robot (AMR) with Computation-in-Memory (CiM) and performing 3D object detection with low energy consumption. Augmented Point Cloud VoxelNet (APCVN) is a network that improves inference accuracy by allowing a slight increase in computational complexity. Multi-Stage Quantization Aware Training (MSQAT) and U-Quantization (UQ) are a learning method and a quantization strategy, respectively, that improve the quantization tolerance of APCVN. Furthermore, quantitative calculations are conducted to estimate ADC energy consumption per inference, operations per second, energy efficiency, memory array area, memory capacity, and latency per inference in the assumed CiM system. Results show that by applying proposed methods, ADC energy consumption per inference is reduced by 8.7% and energy efficiency is improved by 1.6 times while maintaining high inference accuracy.

A protocol for trustworthy EEG decoding with neural networks

Deep learning solutions have rapidly emerged for EEG decoding, achieving state-of-the-art performance on a variety of decoding tasks. Despite their high performance, existing solutions do not fully address the challenge posed by the introduction of many hyperparameters, defining data pre-processing, network architecture, network training, and data augmentation. Automatic hyperparameter search is rarely performed and limited to network-related hyperparameters. Moreover, pipelines are highly sensitive to performance fluctuations due to random initialization, hindering their reliability. Here, we design a comprehensive protocol for EEG decoding that explores the hyperparameters characterizing the entire pipeline and that includes multi-seed initialization for providing robust performance estimates. Our protocol is validated on 9 datasets about motor imagery, P300, SSVEP, including 204 participants and 26 recording sessions, and on different deep learning models. We accompany our protocol with extensive experiments on the main aspects influencing it, such as the number of participants used for hyperparameter search, the split into sequential simpler searches (multi-step search), the use of informed vs. non-informed search algorithms, and the number of random seeds for obtaining stable performance. The best protocol included 2-step hyperparameter search via an informed search algorithm, with the final training and evaluation performed using 10 random initializations. The optimal trade-off between performance and computational time was achieved by using a subset of 3-5 participants for hyperparameter search. Our protocol consistently outperformed baseline state-of-the-art pipelines, widely across datasets and models, and could represent a standard approach for neuroscientists for decoding EEG in a trustworthy and reliable way.

How machine learning can revolutionize building comfort: Accessing the impact of occupancy prediction models on HVAC control system

Ventilation control systems globally constitute among the most significant energy demands in the building industry. Optimizing the energy usage of such systems is imperative for constructing sustainable buildings and is essential for achieving environmental sustainability. Occupancy factors of these buildings, particularly Heating, Ventilation and Air Conditioning (HVAC) optimization, are pivotal for energy optimization. In this work, we apply machine learning approaches to improve the efficiency of the HVAC systems of the Engineering Classroom and Research Building (ENCARB) at Prairie View A&M University. We focus on supporting real-time automated HVAC control through the accurate estimation of HVAC temperature based on occupancy patterns. Therefore, we introduce AIRFLO, an AI-powered Robust Framework for Learning to Optimize HVAC energy consumption. Our framework integrates several occupancy factors obtained from the class schedule to estimate ideal HVAC temperatures. Specifically, we incorporate user nonspecific behavior vis-a-vis energy HVAC service period within the ENCARB Building to gather useful matrices as the basis for the model design. To learn the complex patterns in data, we trained our framework using supervised machine learning. Specifically, we initially trained our framework using an ensemble of multilayer neural networks using our training data and observed the estimation performance on an independent validation set. To learn deeper representations and perform systematic comparative model analysis, we proposed our plan to incorporate both Convolutional Neural Networks (CNNs) and Graph Neural Networks (GNNs). Overall, our proposed approach will offer a comprehensive and scalable framework for optimizing HVAC energy optimization, leading to improved sustainability.

Assessing the Robustness of Test Selection Methods for Deep Neural Networks

Regularly testing deep learning-powered systems on newly collected data is critical to ensure their reliability, robustness, and efficacy in real-world applications. This process is demanding due to the significant time and human effort required for labeling new data. While test selection methods alleviate manual labor by labeling and evaluating only a subset of data while meeting testing criteria, we observe that such methods with reported promising results are simply evaluated, e.g., testing on original test data. The question arises: are they always reliable? In this paper, we explore when and to what extent test selection methods fail. First, we identify potential pitfalls of 11 selection methods based on their construction. Second, we conduct a study to empirically confirm the existence of these pitfalls. Furthermore, we demonstrate how pitfalls can break the reliability of these methods. Concretely, methods for fault detection suffer from data that are: 1) correctly classified but uncertain, or 2) misclassified but confident. Remarkably, the test relative coverage achieved by such methods drops by up to 86.85%. Besides, methods for performance estimation are sensitive to the choice of intermediate-layer output. The effectiveness of such methods can be even worse than random selection when using an inappropriate layer.

Multiple-model-based adaptive state fusion estimation of IWMD-EV using SSUKF

Abstract To improve the preciseness degree and adaptive adjustment ability of the state acquisition system under multiple operating conditions, a vehicle state fusion estimation strategy using strong-tracking suboptimal unscented Kalman filtering (SSUKF) algorithm and adaptive weights is proposed. The vehicle models were established by comprehensively characterizing the multi-parameter mapping relationship under the kinematic, dynamic, and electromechanical coupling driving relationship of in-wheel motor drive electric vehicle (IWMD-EV). To improve estimation performance of vehicle state estimation system in face of nonlinear factors such as noise and interference, a vehicle kinematic-based observer (KO) and a dynamic-based observer (DO) were developed based on the SSUKF algorithm. In addition, to enhance the overall preciseness degree and its adaptive adjustment ability for large-scale road conditions, a state fusion estimation method with adaptive weight is presented by combining redundant information of multiple model observers. The results demonstrated that the proposed method can significantly improve the estimation preciseness degree and the adaptive ability for complex working conditions.

Improving Metrological Performance Estimation of Digital Volume Correlation: Application to X-Ray Computed Tomography

Improving Metrological Performance Estimation of Digital Volume Correlation: Application to X-Ray Computed Tomography

Leveraging Incremental Machine Learning for Reconfigurable Systems Modeling under Dynamic Workloads

Dynamic workload orchestration is one of the main concerns when working with heterogeneous computing infrastructures in the edge-cloud continuum. In this context, FPGA-based computing nodes can take advantage of their improved flexibility, performance and energy efficiency provided that they use proper resource management strategies. In this regard, many state-of-the-art systems rely on proactive power management techniques and task scheduling decisions, which in turn require deep knowledge about the applications to be accelerated and the actual response of the target reconfigurable fabrics when executing them. While acquiring this knowledge at design time was more or less feasible in the past, with applications mostly being static task graphs that did not change at run time, the highly dynamic nature of current workloads in the edge-cloud continuum, where tasks can be deployed on any node and at any time, has removed this possibility. As a result, being able to derive such information at run time to make informed decisions has become a must. This paper presents an infrastructure to build incremental ML models that can be used to obtain run-time power consumption and performance estimations in FPGA-based reconfigurable multi-accelerator systems operating under dynamic workloads. The proposed infrastructure features a novel stop-and-restart resource-aware mechanism to monitor and control the model training and evaluation stages during normal system operation, enabling low-overhead updates in the models to account for either unexpected acceleration requests (i.e., tasks not considered previously by the models) or model drift (e.g., fabric degradation). Experimental results show that the proposed approach induces a maximal additional error of 3.66% compared to a continuous training alternative. Furthermore, the proposed approach incurs only a 4.49% execution time overhead, compared to the 20.91% overhead induced by the continuous training alternative. The proposed modeling strategy enables innovative scheduling approaches in reconfigurable systems. This is exemplified by the conflict-aware scheduler introduced in this work, which achieves up to a 1.35 times speedup in executing the experimental workload. Additionally, the proposed approach demonstrates superior adaptability compared to other methods in the literature, particularly in response to significant changes in workload and to mitigate the effects of model overfitting. The portability of the proposed modeling methodology and monitoring infrastructure is also shown through their application to both Zynq-7000 and Zynq UltraScale+ devices.