Search NASA⌕ Search

SEARCH · Search NASA

Results for “performance modeling”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 55 records · Page 3
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Quantifying leaf symptoms of sorghum charcoal rot in images of field‐grown plants using deep neural networks

Abstract Charcoal rot of sorghum (CRS) is a significant disease affecting sorghum crops, with limited genetic resistance available. The causative agent, Macrophomina phaseolina (Tassi) Goid, is a highly destructive fungal pathogen that targets over 500 plant species globally, including essential staple crops. Utilizing field image data for precise detection and quantification of CRS could greatly assist in the prompt identification and management of affected fields and thereby reduce yield losses. The objective of this work was to implement various machine learning algorithms to evaluate their ability to accurately detect and quantify CRS in red‐green‐blue images of sorghum plants exhibiting symptoms of infection. EfficientNet‐B3 and a fully convolutional network emerged as the top‐performing models for image classification and segmentation tasks, respectively. Among the classification models evaluated, EfficientNet‐B3 demonstrated superior performance, achieving an accuracy of 86.97%, a recall rate of 0.71, and an F1 score of 0.73. Of the segmentation models tested, FCN proved to be the most effective, exhibiting a validation accuracy of 97.76%, a recall rate of 0.68, and an F1 score of 0.66. As the size of the image patches increased, both models’ validation scores increased linearly, and their inference time decreased exponentially. This trend could be attributed to larger patches containing more information, improving model performance, and fewer patches reducing the computational load, thus decreasing inference time. The models, in addition to being immediately useful for breeders and growers of sorghum, advance the domain of automated plant phenotyping and may serve as a foundation for drone‐based or other automated field phenotyping efforts. Additionally, the models presented herein can be accessed through a web‐based application where users can easily analyze their own images.

Gonzalez, Emmanuel M.↗
DOE OSTI2026

Image processing pipeline for AI-driven nanoparticle megalibrary characterization

Recent innovations have made it possible to produce megalibraries, millions of structurally and compositionally distinct nanoparticles on a chip. These megalibraries yield vast volumes of data that are impossible to analyze manually, necessitating the development of automated tools. In previous work, we created a binary classification machine learning model to select quality nanoparticle images for downstream analysis. In this work, we show that adding a custom image processing step before training can produce significantly higher-performing models in a fraction of the time and make them more robust to different image noise levels and microscope acquisition settings. The image processing pipeline proposed here effectively cleans raw nanoparticle images, enhances key features, and allows us to use much lower resolution images and simpler neural network model architectures. These features result in higher performance and significant cost savings. Experiments demonstrate superior performance relative to baseline, including an 18.2% improvement in recall and a 13.1% increase in accuracy. Given the high cost of downstream analysis, it is critical to minimize false positives, and our best-performing model reaches a precision of 95.9% and a weighted F-score of 95.1% on an unseen test set. Additionally, model training time is reduced from hours to less than a minute. We also show that, using this custom image processing pipeline, model performance is significantly improved at lower pixel resolutions compared to downsizing alone. We expect that adopting this pipeline for AI-driven automated nanoparticle characterization will allow researchers to rapidly and accurately analyze much greater volumes of data, thereby accelerating materials discovery.

77 NANOSCIENCE AND NANOTECHNOLOGY↗
DOE OSTI2026

Informing Robust Functional Relationship Benchmarks: An Evaluation of the Temperature Sensitivity of Ecosystem Respiration Across the Arctic-Boreal Region

During land model development, simulated carbon dynamics are often benchmarked against observational data sets to evaluate model performance. Functional relationship benchmarks are the relationship between a driving variable (e.g., temperature) and a response variable (e.g., ecosystem respiration) and are a promising tool for assessing model performance by evaluating modeled sensitivities to changing environmental conditions. However, observed functional relationships can be influenced by choices made during data collection and throughout the benchmarking process, impacting the inferred skill of land models. To avoid misrepresenting a model's true performance, it is necessary to systematically evaluate best practices when constructing functional relationship benchmarks. We developed a set of guidelines for constructing functional relationship benchmarks, considering the choice of data set, number of daily observations, temporal extent, and temporal resolution across Alaska and Canada over a 20-year period from 2001 to 2020. The temperature sensitivity of ecosystem respiration from observations, evaluated through an apparent Q 10 , is highly variable both spatially and as a result of the data processing approach applied in the benchmark formation. When benchmarking 13 models from the Warming Permafrost Model Intercomparison Project (WrPMIP), the range in inferred model skill is substantially impacted by the choices applied in constructing functional relationship benchmarks. The inferred performance of a given model is most sensitive to the number of daily observations and temporal extent, followed by choice of benchmark data set and temporal averaging. Results from this analysis can guide the development of consistent and robust functional relationships for future model evaluation studies.

Poe, Jeralyn [Northern Arizona University, Flagsta↗
DOE OSTI2026

Observation of quantum effects on radiation reaction in strong fields

Radiation reaction, the force experienced by an accelerated charge due to radiation emission, has long been the subject of extensive theoretical and experimental research. Experimental verification of a quantum, strong-field description of radiation reaction is fundamentally important, and has wide-ranging implications for astrophysics, laser-driven particle acceleration, next-generation particle colliders and inverse-Compton photon sources for medical and industrial applications. However, the difficulty of accessing regimes where strong field and quantum effects dominate inhibited previous efforts to observe quantum radiation reaction in charged particle dynamics with high significance. We report the first high significance (> 5σ) observation of strong-field radiation reaction on electron spectra where quantum effects are substantial. We obtain the first, quantitative, strong evidence favouring the quantum-continuous and quantum-stochastic models over the classical model; the quantum models perform comparably. The lower electron energy losses predicted by the quantum models accounts for their improved performance. Model comparison was performed using a novel Bayesian framework which has widespread utility for laser-particle collision experiments, including those utilising conventional accelerators, where some collision parameters cannot be measured directly.

Los, Eva E. [Imperial College London (United Kingd↗
DOE OSTI2024

Accurate and Data‐Efficient Micro X‐ray Diffraction Phase Identification Using Multitask Learning: Application to Hydrothermal Fluids

Traditional analysis of highly distorted micro X‐ray diffraction (μ‐XRD) patterns from hydrothermal fluid environments is a time‐consuming process, often requiring substantial data preprocessing and labeled experimental data. Herein, the potential of deep learning with a multitask learning (MTL) architecture to overcome these limitations is demonstrated. MTL models are trained to identify phase information in μ‐XRD patterns, minimizing the need for labeled experimental data and masking preprocessing steps. Notably, MTL models show superior accuracy compared to binary classification convolutional neural networks. Additionally, introducing a tailored cross‐entropy loss function improves MTL model performance. Most significantly, MTL models tuned to analyze raw and unmasked XRD patterns achieve close performance to models analyzing preprocessed data, with minimal accuracy differences. This work indicates that advanced deep learning architectures like MTL can automate arduous data handling tasks, streamline the analysis of distorted XRD patterns, and reduce the reliance on labor‐intensive experimental datasets.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Adaptive Client Selection in Federated Learning: A Network Anomaly Detection Use Case

Federated Learning (FL) has become a ubiquitous approach for training machine learning models on decentralized data, addressing the myriad privacy concerns inherent in traditional centralized methods. However, the efficiency of FL depends on effective client selection and robust privacy preservation mechanisms. Inadequate client selection may lead to suboptimal model performance, while insufficient privacy measures risk exposing sensitive data. This paper proposes a client selection framework for FL that integrates differential privacy and fault tolerance. Our adaptive approach dynamically adjusts the number of selected clients based on model performance and system constraints, ensuring privacy through calibrated noise addition. We evaluate our method on a network anomaly detection use case using the UNSW-NB15 and ROAD datasets. Results show up to a 7% increase in accuracy and a 25% reduction in training time compared to FedL2P. Moreover, we highlight the trade-offs between privacy budgets and model performance, with higher privacy budgets reducing noise and improving accuracy. Our fault tolerance mechanism, while causing a slight performance drop, enhances robustness to client failures. Statistical validation using Mann-Whitney U tests confirms the significance of these improvements (p < 0.05).

Marfo, William [University of Texas at El Paso,Dep↗
DOE OSTI2026

An improved dataset for predicting mammal infecting viruses from genetic sequence information

There have been several attempts to develop machine learning (ML) models to identify human infecting viruses from their genomic sequences, with varying degrees of success. Direct comparison between models is problematic, because these models are typically trained and evaluated on different datasets with alternative data splitting schemes, features, and model performance metrics. In this paper we present a standardized dataset of mammal infecting and non-infecting viral pathogens, refined from the previous work of Mollentze et al. to include the latest literature evidence, roughly doubling the number of curated host-virus records available to the community, and new host target labels, primate and mammal. The new host labels were included for several reasons, including previous reports that classification performance is better at broader taxonomic ranks and the idea that there may be more data for primate infection that might serve as a suitable proxy for zoonotic potential and avoidance of false positives for human infection due to absence of evidence. On this dataset, we report the performance of eight machine learning models for predicting mammal-infecting viruses from their genomic sequences. We find that randomly assigning cases in our improved dataset to training/testing sets, when compared to the original assignments into training/testing in Mollentze et al., increases the overall average ROC AUC of prediction of human infection from 0.663 ± 0.070 to 0.784 ± 0.013, consistent with the reduction in phylogenetic distance between train and test sets (relative entropy change from 3.00 to 0.08). The broadest host category of mammal infection can be predicted most reliably at 0.850 ± 0.020. We share our improved dataset and code to enable standardized comparisons of machine learning methods to predict human host infections. Overall, we have presented preliminary evidence that classification of virus host infection is more tractable at higher taxonomic ranks, that unsurprisingly reducing the phylogenetic distance between training and test sets can improve predictive performance, that peptide kmer features appear to be harmful to out of sample model performance, and we are left with the question of whether models for virus host prediction can reasonably be expected to perform well in out of sample scenarios given the likelihood that viruses do not share a common ancestor. Consistent with this concern, when the data is resampled such that there is no overlap between viral families in training and test sets (relative entropy > 24), models perform no better than random chance at prediction of human infection regardless of whether kmers are included (ROC AUC 0.50 ± 0.08) or not (ROC AUC 0.50 ± 0.04).

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

Impacts of benchmarking choices on inferred model skill of the Arctic–Boreal terrestrial carbon cycle

Abstract Land surface models require continuous validation against observations to improve and reduce simulation uncertainty. However, inferred model performance can be heavily influenced by subjective choices made in the selection and application of observational data products. A key area often misrepresented by models is the Arctic–Boreal region, which is a potential tipping point region in Earth’s climate system due to large permafrost carbon stocks that are vulnerable to release with climate warming. We use the International Land Model Benchmarking (ILAMB) framework to evaluate how the model skill of TRENDY-v9 models varies based on the choice of observational-based benchmark and how benchmarks are applied in model evaluation. This analysis uses global datasets integrated into ILAMB and new, regionally-specific observational products from the Arctic–Boreal Vulnerability Experiment. Our results cover the overall time period of 1979–2019 and show that model scores can vary substantially depending on the data product applied, with higher model scores indicating better model performance against observations. The lowest model scores occur when benchmarked against regional, compared to global, datasets. We also evaluate observed and modeled functional relationships between ecosystem respiration and air temperature and between gross primary production and precipitation. Here, we find that the magnitude and shape of the responses are strongly impacted by the choice of observational dataset and the approach used to construct the functional relationship benchmark. These results suggest that model evaluation studies could conclude a false sense of model skill if only using a single benchmark data product or if not applying regional data products when performing a regional model analysis. Collectively, our findings highlight the influence of benchmarking choices on model evaluation and point to the need for benchmarking guidelines when assessing model skill.

Poe, Jeralyn (ORCID:0000000318495278)↗
DOE OSTI2025

Intelligent Manufacturing Support: Specialized LLMs for Composite Material Processing and Equipment Operation

Engineering educational curriculum and standards cover many material and manufacturing options. However, engineers and designers are often unfamiliar with certain composite materials or manufacturing techniques. Large language models (LLMs) could potentially bridge the gap. Their capacity to store and retrieve data from large databases provides them with a breadth of knowledge across disciplines. However, their generalized knowledge base can lack targeted, industry-specific knowledge. To this end, we present two LLM-based applications based on the GPT-4 architecture: (1) The Composites Guide: a system that provides expert knowledge on composites material and connects users with research and industry professionals who can provide additional support and (2) The Equipment Assistant: a system that provides guidance for manufacturing tool operation and material characterization. By combining the knowledge of general AI models with industry-specific knowledge, both applications are intended to provide more meaningful information for engineers. In this paper, we discuss the development of the applications and evaluate it through a benchmark and two informal user studies. The benchmark analysis uses the Rouge and Bertscore metrics to evaluate our models’ performance against GPT-4o. The results show that GPT-4o and the proposed models perform similarly or better on the ROUGE and BERTScore metrics. The two user studies supplement this quantitative evaluation by asking experts to provide qualitative and open-ended feedback about our model’s performance on a set of domain-specific questions. The results of both studies highlight a potential for more detailed and specific responses with the Composites Guide and the Equipment Assistant.

Kapoor, Gunnika [Oak Ridge National Laboratory (OR↗
DOE OSTI2025

Exploring Flood Predictability in Taiwan through Coupled Atmospheric–Hydrological and High-Performance Hydrodynamic Models

Effective flood simulation capabilities can tremendously support early warning and disaster prevention. To examine the applicability of a fully physics-based and high-performance flood simulation and forecasting modeling framework for a flood-prone region in Taiwan, we conduct a numerical experiment that couples the Weather Research and Forecasting (WRF) Model, WRF-Hydrological modeling system (WRF-Hydro), and the Two-Dimensional Runoff Inundation Toolkit for Operational Needs (TRITON) to perform integrated rainfall, streamflow, and flood simulations. Furthermore, we first use the coupled WRF and WRF-Hydro (WWH) to predict rainfall and streamflow and then drive TRITON with the predicted streamflow hydrographs to simulate flood depth and inundation area. With the refined spatial resolution and parameterization, this framework can better predict rainfall with reasonable spatial patterns. Although WWH could overestimate the amount of rainfall in some areas, the uncertain rainfall–streamflow predictions produce reasonable flood maps able to pinpoint regions at risk of flooding. In terms of model efficiency, the graphics processing unit–based computation can yield a speed-up factor as high as ∼13 compared to the central processing unit–based computation, promoting the efficacy of the coupled modeling framework in practical real-time flood forecasting.

Coupled models↗
DOE OSTI2024

At Risk Population Estimates for Belarus, Poland and Slovakia with Machine Learning

High-resolution gridded population modeling is crucial for various applications, including disaster response planning, infectious disease spread modeling, climate change impact estimation, policy development, and more. Multiple gridded population datasets have been developed, each tailored to meet specific objectives. Among them, LandScan Global dataset is designed to represent ambient and unwarned population distributions. However, this dataset relies on a statistical approach that requires manual adjustments, making it time consuming and labour intensive. Existing machine learning (ML) methods often train and test at different spatial resolutions, potentially leading to inflated results, and they rely on Census population totals for disaggregation. To address these limitations, in this study we developed population estimates using ML models trained and tested at a consistent 30 arc-second resolution (≈1 square kilometer), specifically using Random Forest (RF) and XGBoost. These models were trained on 2020 datum to predict for 2021 for three countries: Belarus, Poland, and Slovakia. Our findings show that both RF (MAE varies from 5.75 to 13.25) and XGBoost (MAE varies from 8.15 to 23.44) model performance is close to LandScan Global estimates. Furthermore, neither of the models performed the best across all grid cells: the RF model was more effective in areas with lower populations, while XGBoost excelled in more densely populated regions. The proposed approach can be used for countries where the Census data is not available.

Lebakula, Viswadeep [ORNL] (ORCID:0000000152935914↗
DOE OSTI2024

Rapid measurement of soluble xylo-oligomers using near-infrared spectroscopy (NIRS) and multivariate statistics: calibration model development and practical approaches to model optimization

Rapid monitoring of biomass conversion processes using techniques such as near-infrared (NIR) spectroscopy can be substantially quicker and less labor-, resource-, and energy-intensive than conventional measurement techniques such as gas or liquid chromatography (GC or LC) due to the lack of solvents and preparation methods, as well as removing the need to transfer samples to an external lab for analytical evaluation. The purpose of this study was to determine the feasibility of rapid monitoring of a biomass conversion process using NIR spectroscopy combined with multivariate statistical modeling, and to examine the impact of (1) subsetting the samples in the original dataset by process location and (2) reducing the spectral range used in the calibration model on model performance. We develop multivariate calibration models for the concentrations of soluble xylo-oligosaccharides (XOS), monomeric xylose, and total solids at multiple points in a biomass conversion process which produces and then purifies XOS compounds from sugar cane bagasse. A single model using samples from multiple locations in the process stream showed acceptable performance as measured by standard statistical measures. However, compared to the single model, we show that separate models built by segregating the calibration samples according to process location show improved performance. We also show that combining an understanding of the sample spectra with simple multivariate analysis tools can result in a calibration model with a substantially smaller spectral range that provides essentially equal performance to the full-range model. We demonstrate that real-time monitoring of soluble xylo-oligosaccharides (XOS), monomeric xylose, and total solids concentration at multiple points in a process stream using NIR spectroscopy coupled with multivariate statistics is feasible. Segregation of sample populations by process location improves model performance. Models using a reduced spectral range containing the most relevant spectral signatures show very similar performance to the full-range model, reinforcing the importance of performing robust exploratory data analysis before beginning multivariate modeling.

09 BIOMASS FUELS↗
DOE OSTI2024

Chemical signature characterization with hyperspectral imagery: novel deep learning model architectures and physically-motivated data augmentation techniques

The high spectral resolution afforded by Hyperspectral Imaging (HSI) sensors is poised to bring unprecedented advancements to signature characterization applications. Thus far, much of the research in the machine learning field devoted to HSI applications has focused on a few specific tasks like land-use land-cover classification. In land classification tasks, spatial information is very important, and model architectures are often designed to leverage spatial contexts. However, it is unclear how well these spatially-tuned models will translate to tasks where spectral information is critical, like the detection and characterization of chemicals. In this work, we compare spectral models (inputs are 1D spectra) and spatial-spectral models (inputs are 3D cubes) in the context of predicting chemical concentration maps. We find that spatial-spectral models perform the best, though we find a wide range in performance across the different architectures tested. Additionally, we find that model performance is impacted by the availability of training data, particularly in scenarios where the training data doesn't fully capture the true variance of real-world conditions. We find that data augmentation can help mitigate sparse coverage of observed parameter space (e.g., seasonal or geographic variability in ground cover), and present augmentation strategies that are tailored to hyperspectral data.

• Artificial intelligence (AI) / machine learning ↗
DOE OSTI2025

Location generalizability of image-based air quality models

This paper is to be submitted at the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) Computer Vision for Earth Observation workshop. The full paper abstract is below: The ability to rapidly quantify atmospheric pollutants is important both for global emissions monitoring and for mitigating the adverse effects that follow a hazardous chemical release. In the aftermath of a chemical release, imagery is often the only available resource to assess local conditions. Recent work has demonstrated initial success in predicting particulate matter pollution from imagery; however, these results are tied to a specific site and do not generalize to new geographic locations. In this work, we seek to understand how easily deep learning models generalize to new locations in the context of image-based air quality assessments, targeting two distinct tasks: (1) broad measures of particulate matter pollution, and (2) the mass of a given chemical released in hazardous plumes. For the latter, we focus on sulfur dioxide, a toxic aerosol and a major component of particulate matter pollution caused by industrial fossil fuel consumption. To develop a model that operates in the widest possible range of environments, we test different training strategies, including the use of new geolocation foundation models. The best performing models achieve >80% accuracy when evaluating unseen imagery at previously seen sites, but we find significant drops in performance when evaluating imagery from unseen sites, at best 65%. Additionally, we present the public release of the National Parks Air Quality Index Dataset, a new medium-sized dataset that pairs imagery with sensor-based air quality measurements at 15 different national parks.

Byler, Eleanor B. [BATTELLE (PACIFIC NW LAB)]↗
DOE OSTI2024

Leveraging design of experiments to build chemometric models for the quantification of uranium (VI) and HNO3 by Raman spectroscopy

Partial least squares regression (PLSR) and support vector regression (SVR) models were optimized for the quantification of U(VI) (10–320 g L −1 ) and HNO 3 (0.6–6 M) by Raman spectroscopy with optimized calibration sets chosen by optimal design of experiments. The designed approach effectively minimized the number of samples in the calibration set for PLSR and SVR by selecting sample concentrations with a quadratic process model, despite complex confounding and covarying spectral features in the spectra. The top PLS2 model resulted in percent root mean square errors of prediction for U(VI), HNO 3 , and NO 3 − of 3.7%, 3.6%, and 2.9%, respectively. PLS1 models performed similarly despite modeling an analyte with a majority linear response (i.e., uranyl symmetric stretch) and another with more covarying vibrational modes (i.e., HNO 3 ). Partial least squares (PLS) model loadings and regression coefficients were evaluated to better understand the relationship between weaker Raman bands and covarying spectral features. Support vector machine models outperformed PLS1 models, resulting in percent root mean square error of prediction values for U(VI) and HNO 3 of 1.5% and 3.1%, respectively. The optimal nonlinear SVR model was trained using a similar number of samples (11) compared with the PLSR model, even though PLS is a linear modeling approach. The generic D-optimal design presented in this work provides a robust statistical framework for selecting training set samples in disparate two-factor systems. This approach reinforces Raman spectroscopy for the quantification of species relevant to the nuclear fuel cycle and provides a robust chemometric modeling approach to bolster online monitoring in challenging process environments.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Active Learning for Rapid Targeted Synthesis of Compositionally Complex Alloys

The next generation of advanced materials is tending toward increasingly complex compositions. Synthesizing precise composition is time-consuming and becomes exponentially demanding with increasing compositional complexity. An experienced human operator does significantly better than a novice but still struggles to consistently achieve precision when synthesis parameters are coupled. The time to optimize synthesis becomes a barrier to exploring scientifically and technologically exciting compositionally complex materials. This investigation demonstrates an active learning (AL) approach for optimizing physical vapor deposition synthesis of thin-film alloys with up to five principal elements. We compared AL-based on Gaussian process (GP) and random forest (RF) models. The best performing models were able to discover synthesis parameters for a target quinary alloy in 14 iterations. We also demonstrate the capability of these models to be used in transfer learning tasks. RF and GP models trained on lower dimensional systems (i.e., ternary, quarternary) show an immediate improvement in prediction accuracy compared to models trained only on quinary samples. Furthermore, samples that only share a few elements in common with the target composition can be used for model pre-training. We believe that such AL approaches can be widely adapted to significantly accelerate the exploration of compositionally complex materials.

Chemistry↗
DOE OSTI2026

A Benchmarking Framework for Evaluating Large Language Model Capabilities in Nuclear Reactor Safety Applications

Large language models (LLMs) are increasingly capable of answering technical questions, synthesizing domain knowledge, and supporting engineering workflows. For nuclear science and engineering, these capabilities require careful, domain-specific evaluation before they can be credibly incorporated into safety-related activities, regulatory review, or technical decision support. This paper presents preliminary results from benchmarking framework for evaluating LLM capabilities in nuclear contexts. The framework is organized into three evaluation categories: nuclear fundamentals, general dual-use knowledge, and plant specific knowledge. These categories are intended to distinguish general nuclear engineering competence from broader technical reasoning and more context-dependent nuclear knowledge. Initial evaluations focus on nuclear fundamentals using questions representative of the knowledge expected of a nuclear professional engineer. Results indicate that contemporary frontier models perform at a high level and substantially exceed the performance of older model generations, with some models approaching saturation of the current benchmark. These findings suggest both the rapid improvement of LLM capabilities in specialized technical domains and the need for more discriminating evaluation methods. The paper presents the benchmark structure, preliminary model-comparison results, and ongoing work. This work supports development of verifiable, responsible, and safety-conscious methods for assessing AI systems in nuclear engineering applications.

21 - SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLAN↗
DOE OSTI2024

Predictive analytics of selections of russet potatoes

We explore the application of machine learning algorithms specifically to enhance the selection process of Russet potato (Solanum tuberosum L.) clones in breeding trials by predicting their suitability for advancement. This study addresses the challenge of efficiently identifying high-yield, disease-resistant, and climate-resilient potato varieties that meet processing industry standards. Leveraging manually collected data from trials in the state of Oregon, we investigate the potential of a wide variety of state-of-the-art binary classification models. The dataset includes 1086 clones, with data on 38 attributes recorded for each clone, focusing on yield, size, appearance, and frying characteristics, with several control varieties planted consistently across four Oregon regions from 2013 to 2021. We conduct a comprehensive analysis of the dataset that includes preprocessing, feature engineering, and imputation to address missing values. We focus on several key metrics such as accuracy, F1-score, and Matthews correlation coefficient (MCC) for model evaluation. The top-performing models, namely a feedforward neural network classifier (Neural Net), a histogram-based gradient boosting classifier (HGBC), and a support vector machine classifier (SVM), demonstrate consistent and significant results. To further validate our findings, we conducted a simulation study using the aims, data-generating mechanisms, estimands, methods, and performance measures (ADEMP) framework, simulating different data-generating scenarios to assess model robustness and performance through true positive, true negative, false positive, and false negative distributions, area under the receiver operating characteristic curve (AUC-ROC) and MCC. The simulation results highlight that non-linear models like SVM and HGBC consistently show higher AUC-ROC and MCC than logistic regression, thus outperforming the traditional linear model across various distributions, and emphasizing the importance of model selection and tuning in agricultural trials. Variable selection further enhances model performance and identifies influential features in predicting trial outcomes. The findings emphasize the potential of machine learning in streamlining the selection process for potato varieties, offering benefits such as increased efficiency, substantial cost savings, and judicious resource utilization. Our study contributes insights into precision agriculture and showcases the relevance of advanced technologies for informed decision-making in breeding programs.

60 APPLIED LIFE SCIENCES↗