Search NASA⌕ Search

SEARCH · Search NASA

Results for “LSTM 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 91 records · Page 5
All sourcesNASA NTRSDOE OSTI
DOE OSTI2026

L-VISP: LSTM Visualization for Interpretable Symptom Prediction in Patient Cohorts

Symptom modelling in head and neck cancer is challenged by the complexity of heterogeneous patient data, leading to an interest in deep learning approaches. Although Long Short-Term Memory Networks (LSTMs) have shown great results in patient risk prediction, their low interpretability requires data modellers to collaborate with clinical experts to validate the results. We present L-VISP, a human–machine solution that uses visual analytics for LSTM modelling in clinical research. L-VISP uses custom visual encodings to make multiple LSTM variants interpretable, supporting a full range of analysis, from understanding model operations and evaluating performance to interpreting results in a clinical context. We evaluate L-VISP with data modellers and a clinical oncologist and present the takeaways from this multidisciplinary collaboration.

LSTM modeling↗
DOE OSTI2025

Temporal Forecasting of Distributed Temperature Sensing in a Thermal Hydraulic System With Machine Learning and Statistical Models

We benchmark performance of long-short term memory (LSTM) network machine learning model and autoregressive integrated moving average (ARIMA) statistical model in temporal forecasting of distributed temperature sensing (DTS). Data in this study consists of fluid temperature transient measured with two co-located Rayleigh scattering fiber optic sensors (FOS) in a forced convection mixing zone of a thermal tee. We treat each gauge of a FOS as an independent temperature sensor. We first study prediction of DTS time series using Vanilla LSTM and ARIMA models trained on prior history of the same FOS that is used for testing. The results yield maximum absolute percentage error (MaxAPE) and root mean squared percentage error (RMSPE) of 1.58% and 0.06% for ARIMA, and 3.14% and 0.44% for LSTM, respectively. Next, we investigate zero-shot forecasting (ZSF) with LSTM and ARIMA trained on history of the co-located FOS only, which is advantageous when limited training data is available. The ZSF MaxAPE and RMSPE values for ARIMA are comparable to those of the Vanilla use case, while the error values for LSTM increase. We show that in ZSF, performance of LSTM network can be improved by training on most correlated gauges between the two FOS, which are identified by calculating the Pearson correlation coefficient. The improved ZSF MaxAPE and RMSPE for LSTM are 4.4% and 0.33%, respectively. Performance of ZSF LSTM can be further enhanced through transfer learning (TL), where LSTM is re-trained on a subset of the FOS that is the target of forecasting. We show that LSTM pre-trained on correlated dataset and re-trained on 30% of testing target dataset achieves MaxAPE and RMSPE values of 2.32% and 0.28%, respectively.

ARIMA↗
DOE OSTI2022

Differentiable, Learnable, Regionalized Process-Based Models With Multiphysical Outputs can Approach State-Of-The-Art Hydrologic Prediction Accuracy

Predictions of hydrologic variables across the entire water cycle have significant value for water resources management as well as downstream applications such as ecosystem and water quality modeling. Recently, purely data-driven deep learning models like long short-term memory (LSTM) showed seemingly insurmountable performance in modeling rainfall runoff and other geoscientific variables, yet they cannot predict untrained physical variables and remain challenging to interpret. Here, we show that differentiable, learnable, process-based models (called δ models here) can approach the performance level of LSTM for the intensively observed variable (streamflow) with regionalized parameterization. We use a simple hydrologic model HBV as the backbone and use embedded neural networks, which can only be trained in a differentiable programming framework, to parameterize, enhance, or replace the process-based model's modules. Without using an ensemble or post-processor, δ models can obtain a median Nash-Sutcliffe efficiency of 0.732 for 671 basins across the USA for the Daymet forcing data set, compared to 0.748 from a state-of-the-art LSTM model with the same setup. For another forcing data set, the difference is even smaller: 0.715 versus 0.722. Meanwhile, the resulting learnable process-based models can output a full set of untrained variables, for example, soil and groundwater storage, snowpack, evapotranspiration, and baseflow, and can later be constrained by their observations. Both simulated evapotranspiration and fraction of discharge from baseflow agreed decently with alternative estimates. The general framework can work with models with various process complexity and opens up the path for learning physics from big data.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

Surrogate Model of Flexible Research Platform EnergyPlus Models to Enable Sensitivity Analysis

This letter report describes the surrogate models developed from the EnergyPlus model of Oak Ridge National Laboratory’s Flexible Research Platform. Two data-driven black-box models were developed, and the outputs of the surrogate models were compared with the EnergyPlus model. The two models developed are a multilayer perceptron deep learning model, and a long short-term memory (LSTM) neural network model. The three factors for selecting the black-box models are scalability, computation time, and accuracy. A total of 107 input variables were the dominant variables in determining the outputs of building energy consumptions and thermal comfort. A total of 54 output variables were identified as the prediction targets, including the system- and zone-level outputs. The large set of the simulation cases were generated by integrating sensor errors into an emulator based on EnergyPlus and Python EMS, which includes advanced control sequences from ASHRAE Guideline 36-2018: High-Performance Sequences of Operation. The surrogate models were developed based on a set of large-scale simulation runs (i.e., 4,000 runs) on a cloud platform. The comparison analysis shows that the two black-box models had good accuracy for predicting new outputs for sensitivity analysis using the root mean square error metric. As a next step, the developed surrogate models will be used to perform sensitivity analysis for different sensor impacts (e.g., sensor types, sensor locations).

42 ENGINEERING↗
DOE OSTI2025

Techno-economic implications and cost of forecasting errors in solar PV power production using optimized deep learning models

Accurate solar Photovoltaic (PV) power forecasting is important for enhancing both the performance and economic feasibility of PV systems. This study evaluates several deep learning models, including Dense Neural Networks (DNN), Long Short-Term Memory (LSTM), Convolutional Neural Networks (CNN), and a hybrid LSTMCNN model, for predicting PV power production one day in advance. Prior to optimization, the models exhibited relatively high errors, with the best model (DNN) achieving a Root Mean Square Error (RMSE) of 31.13 kW and a coefficient of determination (R 2 ) of 62.15 %. After employing Bayesian optimization, the LSTM-CNN model demonstrated the best performance, with the RMSE reduced to 9.79 kW and R 2 improved to 97.62 %, showcasing significant enhancement in predictive accuracy. Here, the economic evaluation considered three cases: rewards for underestimation (0.08 USD/kWh), no rewards, and penalties for both over-and underestimation (120 % of the utility tariff). In the rewards scenario, the LSTM-CNN model reduced the Levelized Cost of Electricity (LCOE) by 4 %, while in the penalty scenario, a backup diesel generator would have increased the LCOE by 49 %. Additionally, the LSTM-CNN model minimized financial losses, achieving the lowest penalties and maximizing net cash flow compared to other models, demonstrating its overall technical and economic superiority.

Deep learning↗
DOE OSTI2025

S AP F LOWER : an automated tool for sap flow data preprocessing, gap-filling, and analysis using deep learning

Sap flow, a critical process in plant water use and ecosystem water cycles, is often measured using thermal dissipation probes (TDP) due to their ease of installation and continuous data collection. However, sap flow data frequently include noise, outliers, and gaps, creating challenges for analysis and requiring substantial manual processing. We developed S AP F LOWER , a tool that automates data preprocessing, model training, gap-filling, sapwood area scaling and modeling, and water use analysis. It integrates autocleaning, machine learning and deep learning models (e.g. random forest, Gaussian process regression, long short-term memory (LSTM), bidirectional LSTM (BiLSTM)), and efficient workflows to process sap flow data. S AP F LOWER can remove over 90% of noisy data while preserving legitimate variations and achieve high accuracy in gap-filling based on user-determined parameters. Random forest, LSTM, and BiLSTM models reduced root mean square error to 10% or less for long-term gaps. Model training and prediction can be performed efficiently within seconds. S AP F LOWER significantly enhances the efficiency and accessibility of TDP data analysis by automating complex tasks, enabling researchers without programming expertise to employ advanced techniques. Future improvements will focus on species-specific corrections for TDP and support for additional measurement methods. S AP F LOWER is openly available on GitHub (https://github.com/JiaxinWang123/SapFlower) and Zenodo (doi: 10.5281/zenodo.13665919).

ecosystem water balance↗
DOE OSTI2021

Machine Learning Assisted Safety Modeling and Analysis of Advanced Reactors

With the advances in computational power and numerical methods, analysts can now rely on first-principle simulations to predict ultra-fine details in a variety of applications. Advances in machine learning (ML) have produced algorithms that can now learn high-level abstractions via hierarchical models. This project aims to leverage advances in ML techniques and the available high-resolution simulation data to develop a novel modeling and simulation (M\&S) methodology for reactor safety analysis. While application-agnostic ML techniques are available, complex physics constraints need to be incorporated into ML techniques to build ML-based closures for computationally efficient predictive simulations. This project intends to develop a physics-guided data-driven multi-scale methodology for M\&S of advanced reactors. The project focuses on thermal fluid (T/F) phenomena, which play major roles in advanced reactor safety. Specifically, we propose a data-driven coarse-mesh turbulence model based on local flow features for the transient analysis of thermal mixing and stratification in a sodium-cooled fast reactor (SFR). The model has a coarse-mesh setup to ensure computational efficiency, while it is trained by fine-mesh computational fluid dynamics (CFD) data with Reynolds-averaged Navier-Stokes (RANS) turbulence model to ensure accuracy. Three different neural networks are developed and tested for loss-of-flow transients in the hot pool of SFR, i.e. the densely connected convolutional neural network (DCNN), long-short-term-memory network based on proper orthogonal decomposition (POD-LSTM), and the DCNN informed by LSTM (DCNN-LSTM). The performances of these three neural networks are evaluated based on baseline models. The DCNN-LSTM model has been chosen for further hyperparameter optimization. Furthermore, based on a simplified two-dimensional case, uncertainty quantification (UQ) of the developed ML-based closure are investigated with three methods, i.e. Monte Carlo dropout, deep ensemble, and Bayesian neural network. The developed ML-based turbulent viscosity closure relation based on deep ensemble is then integrated into the system analysis module SAM and serves as a term in the conservation equations. Such a SAM-ML based procedure guarantees that the obtained results are consistent with the physical constraints of the thermal-fluid system. The SAM-ML simulation on the same loss-of-flow transient showed comparable accuracy with the CFD simulation but with a much coarser mesh setup. Last but not least, the ML-based closure improvement with the support of higher-fidelity data from large eddy simulation (LES) is discussed. As a first step towards this direction, a baseline LES simulation is performed to obtain comparable data with RANS results. Based on the early results, future investigation on further improving the ML-based closure is discussed. We believe the developed approach that combines scientific machine learning with nuclear system analysis code can benefit the advanced reactor community as more accurate safety analyses will better characterize reactor safety margins and reduce licensing efforts.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2021

Model fusion with physics-guided machine learning: Projection-based reduced-order modeling

The unprecedented amount of data generated from experiments, field observations, and large-scale numerical simulations at a wide range of spatiotemporal scales has enabled the rapid advancement of data-driven and especially deep learning models in the field of fluid mechanics. Although these methods are proven successful for many applications, there is a grand challenge of improving their generalizability. This is particularly essential when data-driven models are employed within outer-loop applications like optimization. In this work, we put forth a physics-guided machine learning (PGML) framework that leverages the interpretable physics-based model with a deep learning model. Leveraging a concatenated neural network design from multi-modal data sources, the PGML framework is capable of enhancing the generalizability of data-driven models and effectively protects against or inform about the inaccurate predictions resulting from extrapolation. We apply the PGML framework as a novel model fusion approach combining the physics-based Galerkin projection model and long- to short-term memory (LSTM) network for parametric model order reduction of fluid flows. We demonstrate the improved generalizability of the PGML framework against a purely data-driven approach through the injection of physics features into intermediate LSTM layers. Our quantitative analysis shows that the overall model uncertainty can be reduced through the PGML approach, especially for test data coming from a distribution different than the training data. Moreover, we demonstrate that our approach can be used as an inverse diagnostic tool providing a confidence score associated with models and observations. The proposed framework also allows for multi-fidelity computing by making use of low-fidelity models in the online deployment of quantified data-driven models.

42 ENGINEERING↗
DOE OSTI2020

Machine learning assisted hybrid models can improve streamflow simulation in diverse catchments across the conterminous US

Incomplete representations of physical processes often lead to structural errors in process-based (PB) hydrologic models. Machine learning (ML) algorithms can reduce streamflow modeling errors but do not enforce physical consistency. As a result, ML algorithms may be unreliable if used to provide future hydroclimate projections where climates and land use patterns are outside the range of training data. Here we test hybrid models built by integrating PB model outputs with a ML algorithm known as Long Short-Term Memory (LSTM) network on their ability to simulate streamflow in 531 catchments representing diverse conditions across the Conterminous United States. Model performance of hybrid models as measured by Nash-Sutcliffe efficiency (NSE) improved relative to standalone PB and LSTM models. More importantly, hybrid models provide highest improvement in catchments where PB models fail completely (i.e., NSE < 0). However, all models performed poorly in catchments with extended low flow periods, suggesting need for additional research.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

Dynamic Data-Driven Multiscale Modeling for Predicting Microstructure and Mechanical Properties of 316L Stainless Steel

We have developed a long short-term memory stacked ensemble (LSTM-SE) surrogate modeling approach that can provide rapid predictions of microstructural evolution and the resultant mechanical properties of American Iron and Steel Institute (AISI) 316L series stainless steel (316LSS) fuel cladding under conditions of varying temperature and radiation dose rate. To acquire training data, we developed and implemented a kinetic Monte Carlo (KMC) model to simulate precipitation kinetics of M 23 C 6 , γ', and G phases within SS316L cladding. Experimentally reported precipitation kinetics of SS316L in literature were linked to the kinetic parameters of the simulated precipitation in our KMC model. The model was then used to simulate microstructure evolution under synthetically generated treatments of varying temperature and radiation dose rate, for periods of up to 3000 hours. Changes in volume fraction, number density, and particle size of precipitates were recorded, and particle area fractions were correlated using statistical methods to develop the surrogate model. Simultaneously, the mechanical properties of the simulated microstructures were evaluated using microstructure-based finite element method (FEM) analysis to determine the elastic modulus, yield stress, ultimate tensile strength, and elongation to failure of the aged microstructures. Using this approach, our surrogate model can predict precipitation behavior within 0.25% volume fraction and mechanical properties within 6% relative error from the values predicted by the KMC and FEM models using 50 training simulations as input. The trained recurrent neural network-based model can return estimations of precipitation kinetics and mechanical properties ~1000 times faster than the physics-based codes. This work demonstrates, as a proof of concept, that reactor material service lifetimes under variable service conditions can be predicted for a statistics-based model from a practicably obtainable dataset.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Dynamic data-driven multiscale modeling for predicting the degradation of a 316L stainless steel nuclear cladding material

Here, we have developed a long short-term memory stacked ensemble (LSTM-SE) surrogate modeling approach that can provide rapid predictions of microstructural evolution and the resultant mechanical properties of American Iron and Steel Institute (AISI) 316L series stainless steel (316LSS) fuel cladding under conditions of varying temperature and radiation dose rate. To acquire training data, we developed and implemented a kinetic Monte Carlo (KMC) model to simulate precipitation kinetics of M 23 C 6 , γ', and G phases within SS316L cladding. Experimentally reported precipitation kinetics of SS316L in literature were linked to the kinetic parameters of the simulated precipitation in our KMC model. The model was then used to simulate microstructure evolution under synthetically generated treatments of varying temperature and radiation dose rate, for periods of up to 3000 hours. Changes in volume fraction, number density, and particle size of precipitates were recorded, and particle area fractions were correlated using statistical methods to develop the surrogate model. Simultaneously, the mechanical properties of the simulated microstructures were evaluated using microstructure-based finite element method (FEM) analysis to determine the elastic modulus, yield stress, ultimate tensile strength, and elongation to failure of the aged microstructures. Using this approach, our surrogate model can predict precipitation behavior within 0.25% volume fraction and mechanical properties within 6% relative error from the values predicted by the KMC and FEM models using 50 training simulations as input. The trained recurrent neural network-based model can return estimations of precipitation kinetics and mechanical properties ~1000 times faster than the physics-based codes. This work demonstrates, as a proof of concept, that reactor material service lifetimes under variable service conditions can be predicted for a statistics-based model from a practicably obtainable dataset.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Increasing Mosquito Abundance Under Global Warming

Mosquitoes are a key virus vector that poses significant health threats globally, affecting 700 million individuals and causing 1 million deaths annually. Accurately predicting mosquito abundance and dispersion remains a challenge. Complex interactions between mosquito dynamics and various environmental factors, notably hydrology, contribute to this challenge. Existing models typically focus on precipitation and temperature and often overlook further impacts of hydrological variables within mosquito modeling. In this study, we developed an artificial intelligence‐based model for mosquito dynamics, explicitly accounting for different hydrological variables, such as precipitation, soil moisture and streamflow. Using Toronto, Canada, as a case study, we identified causal relationships between changes in mosquito populations, hydrological factors, vegetation (e.g., leaf area index), and climate variables (e.g., daylight length, precipitation, and temperature). We embedded these relationships into a Long Short‐Term Memory (LSTM) Neural Network Model capable of accurately detecting mosquito dynamics across annual, seasonal, and monthly time scales. The LSTM is able to explain, on average, approximately 40% of the variance in the observed mosquito abundance data. Using the calibrated model, we predicted that the summer season mosquito abundance would increase by ∼16% and ∼19% under an intermediate greenhouse emission scenario, Shared Socioeconomic Pathway (SSP) 2–4.5, and a high greenhouse emission scenario, SSP5‐8.5, respectively. We expect that this model can serve as a valuable tool and inform science‐based decisions affecting mosquito dynamics and public health. It can also build a foundation for future risk analysis at the regional and larger scales.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2026

Data from: "Towards CONUS-Wide ML-Augmented Conceptually-Interpretable Modeling of Catchment-Scale Precipitation-Storage-Runoff Dynamics"

This data package was generated to support the manuscript “Towards CONUS-Wide Machine Learning-Augmented Conceptually Interpretable Modeling of Catchment-Scale Precipitation-Storage-Runoff Dynamics.” It provides input files, model outputs, plotting data, scripts, notebooks, and documentation used to develop, evaluate, and reproduce Mass-Conserving Perceptron (MCP)-based hydrologic modeling experiments across 513 selected Catchment Attributes and Meteorology for Large-sample Studies in the United States (CAMELS-US) basins. The files are organized by modeling component and analysis purpose, including rainfall–runoff experiments, snow module experiments, coupled hydrologic-snow experiments, Long Short-Term Memory (LSTM) benchmark results, model skill metrics, initialization and epoch records, cell-state normalization files, Akaike Information Criterion (AIC)-based model comparison files, and data used to generate manuscript figures. Tabular files can be opened using standard spreadsheet software or Python/R data-analysis tools. Python scripts, Jupyter notebooks, and selected MATLAB scripts are included for model execution, postprocessing, plotting, and statistical analysis. Quality assurance and quality control were conducted through the source-data selection and modeling workflow. Meteorological forcing, streamflow, and static catchment attributes were derived from the CAMELS-US dataset, and snow water equivalent data were derived from the University of Arizona (UA) Snow Water Equivalent dataset. Selected basins and time periods were screened during the associated research workflow to avoid missing observations or poor-quality cases. Static geospatial features were processed primarily using Quantum Geographic Information System (QGIS) and Geospatial Data Abstraction Library (GDAL) workflows. Additional details are provided in the associated manuscript and documentation.

ESS-DIVE CSV File Formatting Guidelines Reporting ↗
DOE OSTI2021

Hardware-Based Emulator with Deep Learning Model for Building Energy Control and Prediction Based on Occupancy Sensors’ Data

Heating, ventilation, and air conditioning (HVAC) is the largest source of residential energy consumption. Occupancy sensors’ data can be used for HVAC control since it indicates the number of people in the building. HVAC and sensors form a typical cyber-physical system (CPS). In this paper, we aim to build a hardware-based emulation platform to study the occupancy data’s features, which can be further extracted by using machine learning models. In particular, we propose two hardware-based emulators to investigate the use of wired/wireless communication interfaces for occupancy sensor-based building CPS control, and the use of deep learning to predict the building energy consumption with the sensor data. We hypothesize is that the building energy consumption may be predicted by using the occupancy data collected by the sensors, and question what type of prediction model should be used to accurately predict the energy load. Another hypothesis is that an in-lab hardware/software platform could be built to emulate the occupancy sensing process. The machine learning algorithms can then be used to analyze the energy load based on the sensing data. To test the emulator, the occupancy data from the sensors is used to predict energy consumption. The synchronization scheme between sensors and the HVAC server will be discussed. We have built two hardware/software emulation platforms to investigate the sensor/HVAC integration strategies, and used an enhanced deep learning model—which has sequence-to-sequence long short-term memory (Seq2Seq LSTM)—with an attention model to predict the building energy consumption with the preservation of the intrinsic patterns. Because the long-range temporal dependencies are captured, the Seq2Seq models may provide a higher accuracy by using LSTM architectures with encoder and decoder. Meanwhile, LSTMs can capture the temporal and spatial patterns of time series data. The attention model can highlight the most relevant input information in the energy prediction by allocating the attention weights. The communication overhead between the sensors and the HVAC control server can also be alleviated via the attention mechanism, which can automatically ignore the irrelevant information and amplify the relevant information during CNN training. Our experiments and performance analysis show that, compared with the traditional LSTM neural network, the performance of the proposed method has a 30% higher prediction accuracy.

Ye, Zhijing↗
DOE OSTI2025

A Practical Comparison of Data-Driven Prognostics Methods for Energy Systems

This study explores data-driven prognostics for nuclear power plant (NPP) condensers, focusing on tube fouling. We utilized the Asherah nuclear power plant simulator (ANS) to compare four methods: Random Forest (RF), Support Vector Regressor (SVR), Fully Connected Neural Network (FCNN), and Long Short-Term Memory Neural Network (LSTM). By simulating various fouling scenarios in the ANS, we generated data with different degradation rates under transient operations. The models were trained and tested on these data, with performance evaluated visually and numerically including uncertainty assessment. The LSTM model excelled, exhibiting minimal prediction noise and the most accurate remaining useful life estimates across all degradation levels. Its ability to capture long-term dependencies and produce cleaner outputs makes it a strong candidate, although accurate training data across the entire component lifespan are crucial. The RF model emerged as a robust alternative, providing reliable predictions with high confidence. The FCNN and SVR models, while less effective overall, showed potential under specific conditions. FCNN offers a less complex alternative to LSTM and might benefit from larger datasets. SVR excels in precision when the quality of the training data is high. Furthermore, this study highlights the operational benefits of advanced prognostics in the energy sector and emphasizes the need for further research in NPP condenser health management through real-life experiments.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2022

In-Situ Sensing and Dynamics Predictions for Electrothermally-Actuated Soft Robot Limbs

Untethered soft robots that locomote using electrothermally-responsive materials like shape memory alloy (SMA) face challenging design constraints for sensing actuator states. At the same time, modeling of actuator behaviors faces steep challenges, even with available sensor data, due to complex electrical-thermal-mechanical interactions and hysteresis. This article proposes a framework for in-situ sensing and dynamics modeling of actuator states, particularly temperature of SMA wires, which is used to predict robot motions. A planar soft limb is developed, actuated by a pair of SMA coils, that includes compact and robust sensors for temperature and angular deflection. Data from these sensors are used to train a neural network-based on the long short-term memory (LSTM) architecture to model both unidirectional (single SMA) and bidirectional (both SMAs) motion. Predictions from the model demonstrate that data from the temperature sensor, combined with control inputs, allow for dynamics predictions over extraordinarily long open-loop timescales (10 min) with little drift. Prediction errors are on the order of the soft deflection sensor’s accuracy. This architecture allows for compact designs of electrothermally-actuated soft robots that include sensing sufficient for motion predictions, helping to bring these robots into practical application.

Sabelhaus, Andrew P.↗
DOE OSTI2021

CyBERT: Cybersecurity Claim Classification by Fine-Tuning the BERT Language Model

We introduce CyBERT, a cybersecurity feature claims classifier based on bidirectional encoder representations from transformers and a key component in our semi-automated cybersecurity vetting for industrial control systems (ICS). To train CyBERT, we created a corpus of labeled sequences from ICS device documentation collected across a wide range of vendors and devices. This corpus provides the foundation for fine-tuning BERT’s language model, including a prediction-guided relabeling process. We propose an approach to obtain optimal hyperparameters, including the learning rate, the number of dense layers, and their configuration, to increase the accuracy of our classifier. Fine-tuning all hyperparameters of the resulting model led to an increase in classification accuracy from 76% obtained with BertForSequenceClassification’s original architecture to 94.4% obtained with CyBERT. Furthermore, we evaluated CyBERT for the impact of randomness in the initialization, training, and data-sampling phases. CyBERT demonstrated a standard deviation of ±0.6% during validation across 100 random seed values. Finally, we also compared the performance of CyBERT to other well-established language models including GPT2, ULMFiT, and ELMo, as well as neural network models such as CNN, LSTM, and BiLSTM. The results showed that CyBERT outperforms these models on the validation accuracy and the F1 score, validating CyBERT’s robustness and accuracy as a cybersecurity feature claims classifier.

97 MATHEMATICS AND COMPUTING↗