Search NASA⌕ Search

SEARCH · Search NASA

Results for “evaluation datasets”

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 1,189 records · Page 66

Using ARM observations to evaluate NWP results of radiation and cloud fraction

This poster will present the results from: (1) Evaluating NWP model performances in simulating surface cloud radiative forcing measured by effective cloud albedo and total cloud fraction by comparing three major reanalysis datasets (ERAInterim, NCEP/NCAR Reanalysis and NCEP/DOE Reanalysis) against decade-long ARM observations; (2) Examining the relationships between the model biases of the cloud properties and those of nearsurface relative humidity and temperature; (3) Evaluating NWP model performances in simulating vertical profiles of cloud fraction; and, (4) Evaluating NWP model skill scores of predicting cloud fraction.

54 ENVIRONMENTAL SCIENCES↗

Data Analysis & Statistical Methods for Command File Errors

This paper explains current work on modeling for managing the risk of command file errors. It is focused on analyzing actual data from a JPL spaceflight mission to build models for evaluating and predicting error rates as a function of several key variables. We constructed a rich dataset by considering the number of errors, the number of files radiated, including the number commands and blocks in each file, as well as subjective estimates of workload and operational novelty. We have assessed these data using different curve fitting and distribution fitting techniques, such as multiple regression analysis, and maximum likelihood estimation to see how much of the variability in the error rates can be explained with these. We have also used goodness of fit testing strategies and principal component analysis to further assess our data. Finally, we constructed a model of expected error rates based on the what these statistics bore out as critical drivers to the error rate. This model allows project management to evaluate the error rate against a theoretically expected rate as well as anticipate future error rates.

Correlation Analysis↗

Dynamic energy system modeling using hybrid physics-based and machine learning encoder–decoder models

Three model configurations are presented for multi-step time series predictions of the heat absorbed by the water and steam in a thermal power plant. The models predict over horizons of 2, 4, and 6 steps into the future, where each step is a 5-minute increment. The evaluated models are a pure machine learning model, a novel hybrid machine learning and physics-based model, and the hybrid model with an incomplete dataset. The hybrid model deconstructs the machine learning into individual boiler heat absorption units: economizer, water wall, superheater, and reheater. Each configuration uses a gated recurrent unit (GRU) or a GRU-based encoder–decoder as the deep learning architecture. Mean squared error is used to evaluate the models compared to target values. The encoder–decoder architecture is over 11% more accurate than the GRU only models. The hybrid model with the incomplete dataset highlights the importance of the manipulated variables to the system. The hybrid model, compared to the pure machine learning model, is over 10% more accurate on average over 20 iterations of each model. Automatic differentiation is applied to the hybrid model to perform a local sensitivity analysis to identify the most impactful of the 72 manipulated variables on the heat absorbed in the boiler. The models and sensitivity analyses are used in a discussion about optimizing the thermal power plant.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

A report documenting the completion of the Los Alamos National Laboratory portion of the ASC level II milestone ""Visualization on the supercomputing platform

This report provides documentation for the completion of the Los Alamos portion of the ASC Level II 'Visualization on the Supercomputing Platform' milestone. This ASC Level II milestone is a joint milestone between Sandia National Laboratory and Los Alamos National Laboratory. The milestone text is shown in Figure 1 with the Los Alamos portions highlighted in boldfaced text. Visualization and analysis of petascale data is limited by several factors which must be addressed as ACES delivers the Cielo platform. Two primary difficulties are: (1) Performance of interactive rendering, which is the most computationally intensive portion of the visualization process. For terascale platforms, commodity clusters with graphics processors (GPUs) have been used for interactive rendering. For petascale platforms, visualization and rendering may be able to run efficiently on the supercomputer platform itself. (2) I/O bandwidth, which limits how much information can be written to disk. If we simply analyze the sparse information that is saved to disk we miss the opportunity to analyze the rich information produced every timestep by the simulation. For the first issue, we are pursuing in-situ analysis, in which simulations are coupled directly with analysis libraries at runtime. This milestone will evaluate the visualization and rendering performance of current and next generation supercomputers in contrast to GPU-based visualization clusters, and evaluate the perfromance of common analysis libraries coupled with the simulation that analyze and write data to disk during a running simulation. This milestone will explore, evaluate and advance the maturity level of these technologies and their applicability to problems of interest to the ASC program. In conclusion, we improved CPU-based rendering performance by a a factor of 2-10 times on our tests. In addition, we evaluated CPU and CPU-based rendering performance. We encourage production visualization experts to consider using CPU-based rendering solutions when it is appropriate. For example, on remote supercomputers CPU-based rendering can offer a means of viewing data without having to offload the data or geometry onto a CPU-based visualization system. In terms of comparative performance of the CPU and CPU we believe that further optimizations of the performance of both CPU or CPU-based rendering are possible. The simulation community is currently confronting this reality as they work to port their simulations to different hardware architectures. What is interesting about CPU rendering of massive datasets is that for part two decades CPU performance has significantly outperformed CPU-based systems. Based on our advancements, evaluations and explorations we believe that CPU-based rendering has returned as one viable option for the visualization of massive datasets.

97 MATHEMATICS AND COMPUTING↗

TU-C-17A-03: An Integrated Contour Evaluation Software Tool Using Supervised Pattern Recognition for Radiotherapy

Purpose: Radiotherapy (RT) contours delineated either manually or semiautomatically require verification before clinical usage. Manual evaluation is very time consuming. A new integrated software tool using supervised pattern contour recognition was thus developed to facilitate this process. Methods: The contouring tool was developed using an object-oriented programming language C# and application programming interfaces, e.g. visualization toolkit (VTK). The C# language served as the tool design basis. The Accord.Net scientific computing libraries were utilized for the required statistical data processing and pattern recognition, while the VTK was used to build and render 3-D mesh models from critical RT structures in real-time and 360° visualization. Principal component analysis (PCA) was used for system self-updating geometry variations of normal structures based on physician-approved RT contours as a training dataset. The inhouse design of supervised PCA-based contour recognition method was used for automatically evaluating contour normality/abnormality. The function for reporting the contour evaluation results was implemented by using C# and Windows Form Designer. Results: The software input was RT simulation images and RT structures from commercial clinical treatment planning systems. Several abilities were demonstrated: automatic assessment of RT contours, file loading/saving of various modality medical images and RT contours, and generation/visualization of 3-D images and anatomical models. Moreover, it supported the 360° rendering of the RT structures in a multi-slice view, which allows physicians to visually check and edit abnormally contoured structures. Conclusion: This new software integrates the supervised learning framework with image processing and graphical visualization modules for RT contour verification. This tool has great potential for facilitating treatment planning with the assistance of an automatic contour evaluation module in avoiding unnecessary manual verification for physicians/dosimetrists. In addition, its nature as a compact and stand-alone tool allows for future extensibility to include additional functions for physicians’ clinical needs.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Search for a low-energy excess of electron neutrinos in MicroBooNE

The Micro Booster Neutrino Experiment (MicroBooNE) is a Liquid Argon Time Projection Chamber (LArTPC) designed for short-baseline neutrino physics at the Fermi National Accelerator Laboratory. The main physics goal of MicroBooNE is to address the low-energy excess of electron-like events observed by the MiniBooNE experiment and, if confirmed, clarify its nature. The MiniBooNE experiment is a Cherenkov detector and this technology does not allow to distinguish between electrons and single photons in the final state. LArTPC detectors, instead, offer excellent granularity and powerful separation between electrons and photons. For this reason, they represent an ideal technology for the detection of electron neutrino interactions. This thesis presents the first fully-automated electron neutrino selection in a LArTPC. The selection looks for charged-current electron neutrino interactions with no pions and at least one proton in the final state. It is applied on a sub-sample of the data acquired by the detector in the Booster Neutrino Beam, corresponding to $4.34\times10^{19}$~protons-on-target. A validation of the analysis is performed on two orthogonal side-bands, enriched with neutral-current and charged-current muon neutrino interactions, respectively. The uncertainties on the neutrino cross sections, flux, and detector simulation are evaluated. The MicroBooNE detector is placed off-axis with the Neutrinos at the Main Injector (NuMI) beam. An independent dataset of events acquired by triggering on the NuMI beam is employed to measure the significance of the detection of electron neutrinos in the beam using the selection presented here. The sensitivity of the MicroBooNE experiment to the MiniBooNE low-energy excess of electron-like events is evaluated. The efficiency and background-rejection power necessary to achieve $5\sigma$ sensitivity are also quantified.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Maximum likelihood method for cross-correlations with astrophysical sources

We generalize the maximum likelihood-type method used to study cross-correlations between a catalog of candidate astrophysical sources and ultrahigh energy cosmic rays (UHECRs), to allow for differing source luminosities. The new method is applicable to any sparse dataset such as UHE gamma rays or astrophysical neutrinos. Performance of the original and generalized techniques is evaluated in simulations of various scenarios. Applying the new technique to data, we find an excess correlation of about nine events between HiRes UHECRs and known BLLacs, with a 6 Multiplication-Sign 10{sup -5} probability of such a correlation arising by chance.

79 ASTRONOMY AND ASTROPHYSICS↗

ARIADNE – a program estimating covariances in detail for neutron experiments

The python program ARIADNE is a tool developed for evaluators to estimate detailed uncertainties and covariances for experimental data in a consistent and efficient manner. Currently, it is designed to aid in the uncertainty quantification of prompt fission neutron spectra, and was employed to estimate experimental covariances for CIELO and ENDF/B-VIII.0 evaluations. It provides a streamlined way to estimate detailed covariances by (1) implementing uncertainty quantification algorithms specific to the observables, (2) defining input quantities for typically encountered uncertainty sources and correlation shapes, and (3) automatically generating plots of data, uncertainties and correlations, GND formatted XML and plain text output files. Covariances of the same and between different datasets can be estimated, and tools are provided to assemble a database of experimental data and covariances for an evaluation based on ARIADNE outputs. The underlying IPython notebook files can be easily stored, including all assumptions on uncertainties, leading to more reproducible inputs for nuclear data evaluations. Here, the key inputs and outputs are shown along with a representative example for the current version of ARIADNE to illustrate its usability and to open a discussion on how it could address further needs of the nuclear data evaluation community.

97 MATHEMATICS AND COMPUTING↗

SU-F-T-237: The Imaging and Radiation Oncology Core (IROC) Cooperatives Activities Supporting the NCI’s National Clinical Trial Network

Purpose: The Imaging and Radiation Oncology Core (IROC) Cooperative has been active for the past two years supporting the National Clinical Trial Network and the details of that support are reported. Methods: There are six QA centers (Houston, Ohio, Philadelphia-RT, Philadelphia-DI, Rhode Island, St. Louis) providing an integrated RT and DI quality control program in support of the NCI’s clinical trials. The QA Center’s efforts are focused on assuring high quality data for clinical trials designed to improve the clinical outcomes for cancer patients worldwide. This program is administered through five core services: site qualification, trial design support, credentialing, data management, and case review. Results: IROC currently provides core support for 172 NCTN trials with RT, DI and RT/DI components. Many of these trials were legacy trial from the previous cooperative group program. IROC monitors nearly 1800 RT photon and 20 proton institutions. Over 28,000 beams outputs were monitored with 8% of the sites requiring repeat audits due to beam out of criteria. As part of credentialing, 950 QA phantoms have been irradiated, 515 imaging modalities evaluated and almost 4000 credentialing letters have been issued. In just year 2, 5290 RT and 4934 DI patient datasets were received (many using TRIAD) by IROC QA Centers to be prepared for review. During the past 2 years, a total of 6300 RT cases and 19,000 DI image sets were reviewed by IROC technical staff. To date, IROC has published 36 manuscripts. Conclusion: The QA services provided by IROC are numerous and are continually being evaluated for effectiveness, harmonized across all NCTN Groups and administered in an efficient and timely manner to enhance accurate and per protocol trial data submission. These efforts increase each NCTN Group’s ability to derive meaningful outcomes from their clinical trials. This work was supported by DHHS NIH grant 5U24CA180803.

60 APPLIED LIFE SCIENCES↗

HPC Digital Twins for Evaluating Scheduling Policies, Incentive Structures and their Impact on Power and Cooling

Schedulers are critical for optimal resource utilization in high-performance computing. Traditional methods to evaluate sched- ulers are limited to post-deployment analysis, or simulators, which do not model associated infrastructure. In this work, we present the first-of-its-kind integration of scheduling and digital twins in HPC. This enables what-if studies to understand the impact of parameter configurations and scheduling decisions on the physical assets, even before deployment, or regarching changes not easily realizable in production. We (1) provide the first digital twin framework extended with scheduling capabilities, (2) integrate various top-tier HPC systems given their publicly available datasets, (3) implement extensions to integrate external scheduling simulators. Finally, we show how to (4) implement and evaluate incentive structures, as- well-as (5) evaluate machine learning based scheduling, in such novel digital-twin based meta-framework to prototype scheduling. Our work enables what-if scenarios of HPC systems to evaluate sustainability, and the impact on the simulated system.

Maiterth, Matthias [ORNL] (ORCID:000000018698460X)↗

Deep Design Data Portal (D3P) v0.01

The Deep Design Data Portal (D3P) tool was developed to demonstrate how readily accessible data sources, such as building energy model reports for design and baseline energy performance data for projects, can provide the data required for reporting to an industry initiative (AIA 2030 commitment), as well as more detailed data that makes the industry dataset more valuable to all stakeholders, enabling project level analysis and analysis of BEM industry trends. D3P provides an easier and less time-consuming way for firms to auto-extract data from this data source, compared to the current reporting workflows of the firms. The BEM reports are the first of several data sources that D3P could integrate. D3P also provides the ability for firms to review, compare, and evaluate the performance of their projects to not only their portfolio, but also to the larger anonymized industry dataset created each time a project is added to D3P. The intent of D3P is to become part of a data-sharing ecosystem to assist creating large anonymized industry datasets that are accessible to industry.

Regnier, Cynthia [Lawrence Berkeley National Labor↗

SU-F-J-80: Deformable Image Registration for Residual Organ Motion Estimation in Respiratory Gated Treatments with Scanned Carbon Ion Beams

Purpose: At the Centro Nazionale di Adroterapia Oncologica (CNAO, Pavia, Italy) C-ions respiratory gated treatments of patients with abdominal tumours started in 2014. In these cases, the therapeutic dose is delivered around end-exhale. We propose the use of a respiratory motion model to evaluate residual tumour motion. Such a model requires motion fields obtained from deformable image registration (DIR) between 4DCT phases, estimating anatomical motion through interpolation. The aim of this work is to identify the optimal DIR technique to be integrated in the modeling pipeline. Methods: We used 4DCT datasets from 4 patients to test 4 DIR algorithms: Bspline, demons, log-domain and symmetric log domain diffeomorphic demons. We evaluate DIR performance in terms of registration accuracy (RMSE between registered images) and anatomical consistency of the motion field (Jacobian) when registering end-inhale to end-exhale. We subsequently employed the model to estimate the tumour trajectory within the ideal gating window. Results: Within the liver contour, the RMSE is in the range 31–46 HU for the best performing algorithm (Bspline) and 43–145 HU for the worst one (demons). The Jacobians featured zero negative voxels (which indicate singularities in the motion field) for the Bspline fields in 3 of 4 patients, whereas diffeomorphic demons fields showed a non-null number of negative voxels in every case. GTV motion in the gating window measured less than 7 mm for every patient, displaying a predominant superior-inferior (SI) component. Conclusion: The Bspline algorithm allows for acceptable DIR results in the abdominal region, exhibiting the property of anatomical consistency of the computed field. Computed trajectories are in agreement with clinical expectations (small and prevalent SI displacements), since patients lie wearing semi-rigid immobilizing masks. In future, the model could be used for retrospective estimation of organ motion during treatment, as guided by the breathing surrogate signal.

60 APPLIED LIFE SCIENCES↗

Anomaly Detection for the Roman Space Telescope Wide Field Instrument’s Science Data Processing Pipeline

The Roman Space Telescope (RST) Wide Field Instrument (WFI) will be utilizing a preliminary Science Data Processing (SDP) pipeline during its Integration and Test, and to some extent during Operations, to track basic statistics and identify known features such as cosmic rays, snowballs as well as possible anomalies in raw detector data. In our detectors, these anomalies appear as jumps in the ramp of a readout and are classified as cosmic rays if they appear as a streak or snowballs if they’re more circular. The WFI employs an array of 18 H4RG-10 detectors that collect image samples. Each set of raw frames within a non-destructive exposure is packaged by the SDP pipeline into image cubes for each detector. Each cube is a time series of 4096 × 4096 accumulating pixel frames. The preliminary analysis pipeline is used to locate anomalies in these time-series accumulation frames and identify the type of anomaly, either natural phenomena or detector characteristic. To compare different methods, we’ve implemented both heuristic-based and data-driven methods to identify anomalies. For the heuristic-based approach, we identify snowballs and cosmic rays by the size and shape of outlier pixel clusters between consecutive frames. For data driven methods, we evaluated a Convolutional Neural Network (CNN) model, and more traditional methods like Principal Component Analysis (PCA). CNN is a supervised learning/classification method. Thus, we used a labeled dataset of anomalies to perform segmentation of the image and identify anomalies. We used previously identified cosmic rays and snowballs to measure the accuracy and efficiency of the mentioned approaches. In evaluating these methods, we aim to pick the best fit for the SDP pipeline’s anomaly detection in terms of both performance and runtime.

Paul Horton↗

Using Skyline to Analyze Data-Containing Liquid Chromatography, Ion Mobility Spectrometry, and Mass Spectrometry Dimensions

Recent advances in ion mobility spectrometry (IMS) have illustrated its power in determining the structural characteristics of a molecule, especially when coupled with other separations dimensions such as liquid chromatography (LC) and mass spectrometry (MS). However, these three separation techniques together greatly complicate data analyses, making better informatics tools essential for assessing the resulting data. In this manuscript, Skyline was adapted to analyze LC-IMS-CID-MS data from numerous instrument vendor datasets and determine the effect of adding the IMS dimension into the normal LC-MS molecular pipeline. For the initial evaluation, a tryptic digest of bovine serum albumin (BSA) was spiked into a yeast protein digest at seven different concentrations, and Skyline was able to rapidly analyze the MS and CID-MS data for 38 of the BSA peptides. Calibration curves for the precursor and fragment ions were assessed with and without the IMS dimension. In all cases, addition of the IMS dimension removed noise from co-eluting peptides with close m/z values, resulting in calibration curves with greater linearity and lower detection limits. This study presents an important informatics development since to date LC-IMS-CID-MS data from the different instrument vendors is often assessed manually and cannot be analyzed quickly. Because these evaluations require days for the analysis of only a few target molecules in a limited number of samples, it is unfeasible to evaluate hundreds of targets in numerous samples. Thus, this study showcases Skyline’s ability to work with the multidimensional LC-IMS-CID-MS data and provide biological and environmental insights rapidly. .

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Machine learning models inaccurately predict current and future high-latitude C balances

The high-latitude carbon (C) cycle is a key feedback to the global climate system, yet because of system complexity and data limitations, there is currently disagreement over whether the region is a source or sink of C. Recent advances in big data analytics and computing power have popularized the use of machine learning (ML) algorithms to upscale site measurements of ecosystem processes, and in some cases forecast the response of these processes to climate change. Due to data limitations, however, ML model predictions of these processes are almost never validated with independent datasets. To better understand and characterize the limitations of these methods, we develop an approach to independently evaluate ML upscaling and forecasting. We mimic data-driven upscaling and forecasting efforts by applying ML algorithms to different subsets of regional process-model simulation gridcells, and then test ML performance using the remaining gridcells. In this study, we simulate C fluxes and environmental data across Alaska using ecosys, a process-rich terrestrial ecosystem model, and then apply boosted regression tree ML algorithms to training data configurations that mirror and expand upon existing AmeriFLUX eddy-covariance data availability. We first show that a ML model trained using ecosys outputs from currently-available Alaska AmeriFLUX sites incorrectly predicts that Alaska is presently a modeled net C source. Increased spatial coverage of the training dataset improves ML predictions, halving the bias when 240 modeled sites are used instead of 15. However, even this more accurate ML model incorrectly predicts Alaska C fluxes under 21st century climate change because of changes in atmospheric CO 2 , litter inputs, and vegetation composition that have impacts on C fluxes which cannot be inferred from the training data. Our results provide key insights to future C flux upscaling efforts and expose the potential for inaccurate ML upscaling and forecasting of high-latitude C cycle dynamics.

54 ENVIRONMENTAL SCIENCES↗

Systematic Assessment of the Benefits and Caveats in Mining Microbial Post-Translational Modifications from Shotgun Proteomic Data; Response of Shewanella oneidensis to Chromate Exposure

Microbes are known to regulate both gene expression and protein activity through the use of post-translational modifications (PTMs). Common PTMs involved in cellular signaling and gene control include methylations, acetylations, and phosphorylations; whereas oxidations have been implicated as an indicator for stress. Shewanella oneidensis MR-1 is a gram-negative bacterium that demonstrates both respiratory versatility and the ability to sense and adapt to diverse environmental conditions. The dataset used in this study consisted of tandem mass spectra derived from mid-log phase aerobic cultures of S. oneidensis shocked either with or without 1 mM chromate [Cr(VI)]. In this study, three algorithms (DBDigger, Sequest, and InsPecT) were evaluated for their ability to scrutinize shotgun proteomic data for evidence of PTMs. The use of conservative scoring filters for peptides or proteins versus creating a sub-database first from a non-modification search was evaluated with DBDigger. The use of higher scoring filters for peptide identifications was found to result in optimal identifications of PTM peptides with a 2% false discovery rate (FDR) for the total dataset using the DBDigger algorithm. However, the FDR climbs to about 50% when considering PTM peptides only. Sequest was evaluated as a method for confirming PTM peptides putatively identified using DBDigger; however, there was a low identification rate (~25%) for the searched spectra. InsPecT was found to have a lower FDR (~9%) than DBDigger for PTM peptides. Comparisons between InsPecT and DBDigger were made with respect to both the FDR and PTM peptide identifications. As a demonstration of this approach, a number of S. oneidensis chemotaxis proteins as well as low-abundance signal transduction proteins were identified as being post-translationally modified in response to chromate challenge.

59 BASIC BIOLOGICAL SCIENCES↗

A leaf-level spectral library to support high-throughput plant phenotyping: predictive accuracy and model transfer

Abstract Leaf-level hyperspectral reflectance has become an effective tool for high-throughput phenotyping of plant leaf traits due to its rapid, low-cost, multi-sensing, and non-destructive nature. However, collecting samples for model calibration can still be expensive, and models show poor transferability among different datasets. This study had three specific objectives: first, to assemble a large library of leaf hyperspectral data (n=2460) from maize and sorghum; second, to evaluate two machine-learning approaches to estimate nine leaf properties (chlorophyll, thickness, water content, nitrogen, phosphorus, potassium, calcium, magnesium, and sulfur); and third, to investigate the usefulness of this spectral library for predicting external datasets (n=445) including soybean and camelina using extra-weighted spiking. Internal cross-validation showed satisfactory performance of the spectral library to estimate all nine traits (mean R2=0.688), with partial least-squares regression outperforming deep neural network models. Models calibrated solely using the spectral library showed degraded performance on external datasets (mean R2=0.159 for camelina, 0.337 for soybean). Models improved significantly when a small portion of external samples (n=20) was added to the library via extra-weighted spiking (mean R2=0.574 for camelina, 0.536 for soybean). The leaf-level spectral library greatly benefits plant physiological and biochemical phenotyping, whilst extra-weight spiking improves model transferability and extends its utility.

59 BASIC BIOLOGICAL SCIENCES↗

A novel data gaps filling method for solar PV output forecasting

This study proposes a modified gaps filling method, expanding the column mean imputation method and evaluated using randomly generated missing values comprising 5%, 10%, 15%, and 20% of the original data on power output. The XGBoost algorithm was implemented as a forecasting model using the original and processed datasets and two sources of solar radiation data, namely, Shortwave Radiation (SWR) from Advanced Himawari Imager 8 (AHI-8) and Surface Solar Radiation Downward (SSRD) from ERA5 global reanalysis data. Further, the accuracy of the two sets of forecasted power output was evaluated using Root Mean Square Error (RMSE) and Mean Absolute Error (MAE). Results show that by applying the proposed gap filling method and using SWR in forecasting solar photovoltaic (PV) output, the improvement in the RMSE and MAE values range from 12.52% to 24.30% and from 21.10% to 31.31%, respectively. Meanwhile, using SSRD, the improvement in the RMSE values range from 14.01% to 28.54% and MAE values from 22.39% to 35.53%. To further evaluate the accuracy of the proposed gap-filling method, the proposed method could be validated using different datasets and other forecasting methods. Future studies could also consider applying the said method to datasets with data gaps higher than 20%.

Energy & Fuels↗