SEARCH · Engineering Papers

Results for “Machine learning prediction”

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.

SourceResearch connectionFrom yearThrough year

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

At least 757 records · Page 42

All sources NASA NTRS DOE OSTI

DOE OSTI2026

The Role of Nuclear Data Sensitivities in Prompt α-Eigenvalue Predictions of Delayed Critical Benchmarks

Alpha (α) eigenvalues, which describe the logarithmic time derivative of the neutron population in a multiplying system, are integral to time-dependent behavior and diagnostic applications. However, uncertainties in the evaluated nuclear data can significantly impact the accuracy of transport simulations for such quantities. This work explores the use of machine learning models to predict two key outputs, α-eigenvalues and keff bias, using input features derived from α-eigenvalue sensitivities to nuclear data. The criticality safety benchmark models used in this study come from the International Handbook of Evaluated Criticality Safety Benchmark Experiments. Three models, random forest, XGBoost, and NGBoost, are trained on both energy-resolved and energy-summed α sensitivities. For the α-eigenvalue bias prediction, NGBoost achieved the highest R 2 (0.9476) using energy-resolved features, while XGBoost performed best using summed sensitivities. In contrast, when predicting the keff bias, all the models showed moderate predictive capability (best R 2 ≈ 0.72), as the mapping from the static α-sensitivities to the static keff bias was less direct. SHAP (SHapley Additive exPlanations) analysis was used to interpret the model predictions. Across both prediction tasks, the features associated with neutron capture [H-1 (n, γ)], uranium scattering reactions (such as 235 U elastic/inelastic), and actinide capture/fission reactions (such as 239 Pu and 234 U) were consistently identified as the most impactful. This highlights the key role of specific nuclear reactions and energy ranges in shaping both time-dependent and steady-state criticality behavior. These results demonstrated that α-sensitivities, despite being computed for time-dependent metrics, can provide valuable insights for predicting both α-eigenvalues and the keff bias. Moreover, machine learning models offer a promising pathway for uncovering important nuclear data dependencies and guiding future data evaluation efforts.

Nuclear data↗

DOE OSTI2020

Automation of Active Space Selection for Multireference Methods via Machine Learning on Chemical Bond Dissociation

Predicting and understanding the chemical bond is one of the major challenges of computational quantum chemistry. Kohn–Sham density functional theory (KS-DFT) is the most common method, but approximate density functionals may not be able to describe systems where multiple electronic configurations are equally important. Multiconfigurational wave functions, on the other hand, can provide a detailed understanding of the electronic structures and chemical bonds of such systems. In the complete active space self-consistent field (CASSCF) method, one performs a full configuration interaction calculation in an active space consisting of active electrons and active orbitals. However, CASSCF and its variants require the selection of these active spaces. This choice is not black box; it requires significant experience and testing by the user, and thus active space methods are not considered particularly user-friendly and are employed only by a minority of quantum chemists. Our goal is to popularize these methods by making it easier to make good active space choices. Here, we present a machine learning protocol that performs an automated selection of active spaces for chemical bond dissociation calculations of main group diatomic molecules. The protocol shows high prediction performance for a given target system as long as a properly correlated system is chosen for training. Good active spaces are correctly predicted with a considerably better success rate than random guess (larger than 80% precision for most systems studied). Our automated machine learning protocol shows that a “black-box” mode is possible for facilitating and accelerating the large-scale calculations on multireference systems where single-reference methods such as KS-DFT cannot be applied.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2025

Effect of particle size and moisture on flow performance of loblolly pine anatomical fractions: Experimental findings and model predictions

The rising energy demand has highlighted biomass as a promising next-generation energy source. However, commercializing biomass-derived energy faces challenges, particularly in handling biomass feedstock. Factors like particle size, shape, moisture content, and surface roughness significantly impact biomass flowability. This study addresses a crucial knowledge gap by examining the effects of particle size and moisture content on the flow behavior and shear properties of different anatomical fractions of loblolly pine (Pinus taeda). The bulk shear behavior was examined using a Schulze ring shear tester, while flow performance was tested through gravity-driven flow experiments in a variable wedge-shape hopper. Results were incorporated into empirical and machine learning-based flow prediction models to evaluate their accuracy and limitations. The study found that samples with higher moisture content show higher unconfined yield strength. The critical arching distance increased with particle size, e.g., from approximately 13 and 33 mm for 2- and 6-mm whole chips, respectively at a 32-degree inclination angle. Conversely, the flow rate decreased for a given hopper opening as particle size increased. For instance, at a 60-mm hopper opening and a 32-degree inclination angle, the mass flow rates for 2- and 6-mm whole chips were 7.83 and 6.42 tonne/h, respectively. The empirical model consistently overpredicted the mass flow rate for all anatomical fractions, while the machine learning model more accurately predicted the central tendency of flow rate but was insensitive to varying tissue proportions. These novel findings provide comprehensive characterization of anatomical fractions, reveal significant combined effects of particle size and moisture content on biomass flow behavior, and demonstrate a better predictive accuracy of a machine learning model, all of which are useful for optimizing material handling strategies and biomass utilization technologies in the industry.

09 - BIOMASS FUELS↗

DOE OSTI2021

A Comparison of Machine Learning Methods to Forecast Tropospheric Ozone Levels in Delhi

Ground-level ozone is a pollutant that is harmful to urban populations, particularly in developing countries where it is present in significant quantities. It greatly increases the risk of heart and lung diseases and harms agricultural crops. This study hypothesized that, as a secondary pollutant, ground-level ozone is amenable to 24 h forecasting based on measurements of weather conditions and primary pollutants such as nitrogen oxides and volatile organic compounds. We developed software to analyze hourly records of 12 air pollutants and 5 weather variables over the course of one year in Delhi, India. To determine the best predictive model, eight machine learning algorithms were tuned, trained, tested, and compared using cross-validation with hourly data for a full year. The algorithms, ranked by R2 values, were XGBoost (0.61), Random Forest (0.61), K-Nearest Neighbor Regression (0.55), Support Vector Regression (0.48), Decision Trees (0.43), AdaBoost (0.39), and linear regression (0.39). When trained by separate seasons across five years, the predictive capabilities of all models increased, with a maximum R 2 of 0.75 during winter. Bidirectional Long Short-Term Memory was the least accurate model for annual training, but had some of the best predictions for seasonal training. Out of five air quality index categories, the XGBoost model was able to predict the correct category 24 h in advance 90% of the time when trained with full-year data. Separated by season, winter is considerably more predictable (97.3%), followed by post-monsoon (92.8%), monsoon (90.3%), and summer (88.9%). These results show the importance of training machine learning methods with season-specific data sets and comparing a large number of methods for specific applications.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2021

Machine learned features from density of states for accurate adsorption energy prediction

Materials databases generated by high-throughput computational screening, typically using density functional theory (DFT), have become valuable resources for discovering new heterogeneous catalysts, though the computational cost associated with generating them presents a crucial roadblock. Hence there is a significant demand for developing descriptors or features, in lieu of DFT, to accurately predict catalytic properties, such as adsorption energies. Here, we demonstrate an approach to predict energies using a convolutional neural network-based machine learning model to automatically obtain key features from the electronic density of states (DOS). The model, DOSnet, is evaluated for a diverse set of adsorbates and surfaces, yielding a mean absolute error on the order of 0.1 eV. In addition, DOSnet can provide physically meaningful predictions and insights by predicting responses to external perturbations to the electronic structure without additional DFT calculations, paving the way for the accelerated discovery of materials and catalysts by exploration of the electronic space.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2024

Capturing Infrastructure Interdependencies for Power Outages Prediction During Extreme Events

As extreme weather events such as hurricanes, severe thunderstorms, and floods grow in frequency and intensity, the disruption of power grid systems poses significant challenges, including widespread electrical outages, economic losses, and threats to public safety. This paper presents a forward-looking approach that leverages geographical graph-based machine learning models to predict county-level maximum power outages during such events. By capturing the intricate interdependencies within power system networks, our approach aims to provide precise and actionable predictions that can optimize emergency response efforts and enhance grid resilience. Through the integration of real-world data, including hurricane advisories and power outage records, we have trained and benchmarked multiple machine learning models, demonstrating the feasibility and potential of this method. While our initial results are promising, this paper also charts a course for advancing these models, addressing the remaining challenges, and ultimately transforming how we anticipate and respond to the impacts of extreme weather on power systems.

Lee, Sangkeun (Matt) [ORNL] (ORCID:000000021317511↗

DOE OSTI2021

Accurate prediction of mega-electron-volt electron beam properties from UED using machine learning

To harness the full potential of the ultrafast electron diffraction (UED) and microscopy (UEM), we must know accurately the electron beam properties, such as emittance, energy spread, spatial-pointing jitter, and shot-to-shot energy fluctuation. Owing to the inherent fluctuations in UED/UEM instruments, obtaining such detailed knowledge requires real-time characterization of the beam properties for each electron bunch. While diagnostics of these properties exist, they are often invasive, and many of them cannot operate at a high repetition rate. Here, we present a technique to overcome such limitations. Employing a machine learning (ML) strategy, we can accurately predict electron beam properties for every shot using only parameters that are easily recorded at high repetition rate by the detector while the experiments are ongoing, by training a model on a small set of fully diagnosed bunches. Applying ML as real-time noninvasive diagnostics could enable some new capabilities, e.g., online optimization of the long-term stability and fine single-shot quality of the electron beam, filtering the events and making online corrections of the data for time-resolved UED, otherwise impossible. This opens the possibility of fully realizing the potential of high repetition rate UED and UEM for life science and condensed matter physics applications.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

DOE OSTI2020

Machine learning study of magnetism in uranium-based compounds

Actinide and lanthanide-based materials display exotic properties that originate from the presence of itinerant or localized f electrons and include unconventional superconductivity and magnetism, hidden order, and heavy-fermion behavior. Due to the strongly correlated nature of the 5f electrons, magnetic properties of these compounds depend sensitively on applied magnetic field and pressure, as well as on chemical doping. However, precise connection between the structure and magnetism in actinide-based materials is currently unclear. In this investigation, we established such structure-property links by assembling and mining two datasets that aggregate, respectively, the results of high-throughput density functional theory simulations and experimental measurements for the families of uranium- and neptunium-based binary compounds. Various regression algorithms were utilized to identify correlations among accessible attributes (features or descriptors) of the material systems and predict their cation magnetic moments and general forms of magnetic ordering. Descriptors representing compound structural parameters and cation f-subshell occupation numbers were identified as most important for accurate predictions. The best machine learning model developed employs the random forest regression algorithm. It can predict both spin and orbit moment size with root-mean-square error of 0.17 μ B and 0.19 μ B , respectively. Lastly, the random forest classification algorithm is used to predict the ordering (paramagnetic, ferromagnetic, and antiferromagnetic) of such systems with 76% accuracy.

36 MATERIALS SCIENCE↗

DOE OSTI2021

Improved departure from nucleate boiling prediction in rod bundles using a physics-informed machine learning-aided framework

The critical heat flux (CHF) corresponding to the departure from nucleate boiling (DNB) crisis is a regulatory limit for the licensing of pressurized water reactors (PWRs) worldwide. Despite the abundance of predictive tools available to the reactor thermal-hydraulics community, the path for an accurate CHF model remains elusive. This work approaches the prediction of DNB through a physics-informed machine learning-aided framework (PIMLAF) with the objective of achieving superior predictive capabilities for a rod bundle. In view of the limitations in existing macro-scale physics-driven tools, an improved mechanistic model is first proposed, leveraging key concepts in the liquid sublayer dryout and bubble crowding mechanisms. Furthermore, the proposed mechanistic model is able to predict DNB in different heater geometries for a broad range of flow conditions without the need for recalibration. This model is then incorporated as the physics-informed component of the hybrid framework PIMLAF, which takes advantage of established understanding in the field (i.e., domain knowledge [DK]) and uses machine learning (ML) to capture undiscovered information from the mismatch between the actual and DK-predicted output. Two bundle-related case studies using the PWR subchannel and bundle tests (PSBT) database are carried out to illustrate the PIMLAF’s improved performance over traditional approaches for both interpolation and extrapolation purposes. In light of the PIMLAF’s promising potential to reduce prediction error, reactor vendors are encouraged to leverage their in-house experimental efforts and apply the hybrid framework to potentially achieve margin reductions in the minimum DNB ratio (MDNBR) for the designs of interest.

42 ENGINEERING↗

DOE OSTI2025

Multiscale and Machine Learning Modeling for Process-informed Microstructure Prediction in Additively Manufactured Materials using MALAMUTE

The Advanced Materials and Manufacturing Technologies (AMMT) program under the Department of Energy Office of Nuclear Energy aims to develop and qualify additively manufactured materials for nuclear applications. One key challenge to this is the microstructural variability observed in the additively manufactured products and their impact on the properties and performance of the material in extreme environments. AMMT is using a combination of high-throughput experimental and modeling techniques to accelerate qualification. Conventionally, in-situ and ex-situ characterizations and testing are performed to correlate different aspects of the additive manufacturing process to the final product and its performance. However, adopting a trial-and-error approach to experimentally evaluate the vast range of process parameters required to capture microstructural variability is cost-prohibitive. Modeling and simulation provide a comparatively inexpensive way to understand and correlate the microstructural evolution to the processing conditions. The modeling and simulation work-packages within the AMMT program aims to use physics-based and machine learning models to develop a digital twin for additive manufacturing that can correlate the process conditions to the final product and establish a process-structure-property-performance (PSPP) correlation. The melting and subsequent solidification that occurs during the additive process is a complex phenomenon that requires multiscale multiphysics analysis. This work package focuses on understanding the role of process variabilities on the unique microstructural characteristics of additively manufactured materials. Microstructural features at the subgrain level, such as compositional micro-heterogeneity and dislocation cells, are of particular interest here since they can influence the creep properties and radiation performance. Idaho National Laboratory’s Multiphysics Object-Oriented Simulation Environment (MOOSE), specifically the MOOSE Application Library for Advanced Manufacturing UTilitiEs (MALAMUTE) software, provides an ideal platform for developing the multiphysics multiscale model to explore the intricacies of the microstructural evolution during the AM processes within a single framework. Furthermore, given that such full-fidelity simulations can be computationally intensive, reduced order models are necessary to explore the PSPP space for additively manufactured materials in an efficient, reliable, and cost-effective way. This work focuses on capturing the microstructural variabilities at the subgrain level that are often missing in the part-scale models. In fiscal year 2025, we significantly advanced upon our work in the last fiscal year, in terms of the predictive capabilities of the physics-based and ML models, by adding the capabilities to capture subgrain-level micro-segregation during solidification using phase-field model and to predict the time-dependent dynamics of the AM process through the MOGPAR model. The alloy solidification model in MOOSE incorporates the thermodynamic properties and free energy relevant to 316 stainless steel. The model demonstrates the Cr and Ni segregation that occurs during solidification, including that the rate of solidification. The microstructural evolution model is connected to the process conditions via the surrogate model developed in this work. This enables predictions of the final microstructure in conjunctions with the manufacturing process. This work supports AMMT's rapid qualification goals by laying the foundation for an efficient and cost-effective model establishing the PSPP correlation for AM. The generated microstructures and predicted micro-segregation can be used by other work packages under AMMT to evaluate the properties and environmental response of the material at the mesoscale. Thus, this work helps to identify the key microstructural features at the subgrain level that are significant in property and performance predictions of additively manufactured components. This work will also provide inputs to the large-scale process variability models to reevaluate and validate assumptions and simplifications made in the part-scale models. Furthermore, through active learning this work can help identify the data need from both modeling and experimental sides for development of a robust digital twin for additive manufacturing and accelerate the AMMT's qualification efforts.

36 - MATERIALS SCIENCE↗

DOE OSTI2025

Heuristic Evaluation Methods Applied to a Predictive Maintenance Chatbot

The need for an accessible iterative approach for evaluating prospective artificial intelligence (AI)/ML based technologies in the nuclear industry is needed, given the nature of algorithms and rapid advancements. This paper explores existing heuristic design principles for user-centered design and evaluates them based on their relevancy and usefulness for evaluating AI/ ML based technologies. Researchers at the Idaho National Laboratory (INL) have developed a machine learning software application called VIsualization for PrEdictive maintenance Recommendation (VIPER), which is used to help users understand and engage with the tool to learn more about work orders, data used, predictive maintenance, and machine learning (ML) algorithms. Early user research studies used to access VIPER’s technology readiness level have occurred; however, there is room for further improvement of the software through heuristic evaluations along with other methods and user testing. This work describes the applicability of heuristic evaluation methods and cognitive walkthroughs to help ensure human readiness for prospective AI/ ML based applications, using VIPER as a candidate use case. This work supports industry in ensuring that prospective AI/ML based technologies are usable and useful for plant personnel at nuclear power plants, ultimately leading to their safe, reliable, and efficient use.

99 - GENERAL AND MISCELLANEOUS↗

DOE OSTI2025

Heuristic Evaluation Methods Applied to a Predictive Maintenance Chatbot

The need for an accessible iterative approach for evaluating prospective artificial intelligence (AI)/ML based technologies in the nuclear industry is needed, given the nature of algorithms and rapid advancements. This paper explores existing heuristic design principles for user-centered design and evaluates them based on their relevancy and usefulness for evaluating AI/ ML based technologies. Researchers at the Idaho National Laboratory (INL) have developed a machine learning software application called VIsualization for PrEdictive maintenance Recommendation (VIPER), which is used to help users understand and engage with the tool to learn more about work orders, data used, predictive maintenance, and machine learning (ML) algorithms. Early user research studies used to access VIPER?s technology readiness level have occurred; however, there is room for further improvement of the software through heuristic evaluations along with other methods and user testing. This work describes the applicability of heuristic evaluation methods and cognitive walkthroughs to help ensure human readiness for prospective AI/ ML based applications, using VIPER as a candidate use case. This work supports industry in ensuring that prospective AI/ML based technologies are usable and useful for plant personnel at nuclear power plants, ultimately leading to their safe, reliable, and efficient use. PowerPoint for conference that was reviewed in PRS and LRS PRS/CON-25-05379 and INL/CON-25-82946

99 - GENERAL AND MISCELLANEOUS↗

DOE OSTI2026

Computational methods in solution-based plastics purification

Plastic waste can be recycled into resins with near-virgin properties by solution-based purification processes that selectively dissolve polymers, remove contaminants, or detach printing residues. Here, in this review, we examine computational methods for predicting the behavior governing solution-based plastic purification, motivated by the vast polymer–solvent–contaminant compositional space. We discuss thermodynamic and machine learning methods for predicting polymer–solvent and polymer–contaminant interaction and review physics-based molecular dynamics simulations that resolve molecular-scale phenomena within polymer matrices inaccessible to screening methods. We highlight how these methods have informed experimental design for dissolution-based recycling and solvent-based contaminant removal. Finally, we discuss the prospective role of agentic AI in integrating these computational tools with real-time sorting data to adapt purification conditions to the compositional variability of real post-consumer feedstocks. This review charts a path toward computationally guided solution-based purification workflows that can respond to the complexity inherent in plastic waste streams.

Altamimi, Ali [Univ. of Wisconsin, Madison, WI (Un↗

DOE OSTI2026

Augmenting Molecular Graphs with Geometries via Machine Learning Interatomic Potentials

Accurate molecular property predictions require 3D geometries, which are typically obtained using expensive methods such as density functional theory (DFT). Here, we attempt to obtain molecular geometries by relying solely on machine learning interatomic potential (MLIP) models. To this end, we first curate a large-scale molecular relaxation dataset comprising 3.5 million molecules and 300 million snapshots. Then MLIP pre-trained models are trained with supervised learning to predict energy and forces given 3D molecular structures. Once trained, we show that the pre-trained models can be used in different ways to obtain geometries either explicitly or implicitly. First, it can be used to obtain approximate low-energy 3D geometries via geometry optimization. While these geometries do not consistently reach DFT-level chemical accuracy or convergence, they can still improve downstream performance compared to non-relaxed structures. To mitigate potential biases and enhance downstream predictions, we introduce geometry fine-tuning based on the relaxed 3D geometries. Second, the pre-trained models can be directly fine-tuned for property prediction when ground truth 3D geometries are available. Our results demonstrate that MLIP pre-trained models trained on relaxation data can learn transferable molecular representations to improve downstream molecular property prediction and can provide practically valuable but approximate molecular geometries that benefit property predictions. Our code is publicly available at: https://github.com/divelab/AIRS/.

Fu, Cong [Texas A & M Univ., College Station, TX (↗

DOE OSTI2022

Sensitivity Analysis of Wind and Turbulence Predictions With Mesoscale‐Coupled Large Eddy Simulations Using Ensemble Machine Learning

Abstract Coupling between mesoscale models and large‐eddy simulation (LES) models is increasingly used to more realistically represent the wide range of scales of atmospheric motions affecting boundary layer winds and turbulence that need to be simulated accurately for applications such as wind energy. However, such mesoscale‐to‐microscale coupled modeling frameworks are potentially affected by a large number of uncertain closure parameters. Here, we investigate the sensitivity associated with six closure parameters related to a 1.5‐order subgrid‐scale turbulence closure for an ensemble of mesoscale‐coupled LES. The simulations are performed using the Weather Research and Forecasting model nested from horizontal resolutions of greater than a kilometer down to tens of meters. Closure parameters are varied to generate perturbed parameter ensembles for two case studies of highly sheared, convective boundary layers observed in the Columbia Basin of Oregon and Washington during the Second Wind Forecast Improvement Project. Machine learning algorithms are used to explore the sensitivity of LES predictions, considering the effects of the perturbed physical parameters alongside categorical factors such as the case study identity, measurement location, and LES resolution. For the conditions we examine, a single parameter, the eddy viscosity coefficient, is the dominant source of parametric sensitivity and its importance is comparable to the categorical factors for several of the simulation response variables we examine.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2023

Accurate Prediction of Adiabatic Ionization Potentials of Organic Molecules using Quantum Chemistry Assisted Machine Learning

In previous work (Dandu et al., J. Phys. Chem. A, 2022, 126, 4528–4536), we were successful in predicting accurate atomization energies of organic molecules using machine learning (ML) models, obtaining an accuracy as low as 0.1 kcal/mol compared to the G4MP2 method. In this work, we extend the use of these ML models to adiabatic ionization potentials on data sets of energies generated using quantum chemical calculations. Atomic specific corrections that were found to improve atomization energies from quantum chemical calculations have also been used in this study to improve ionization potentials. Here, the quantum chemical calculations were performed on 3405 molecules containing eight or fewer non-hydrogen atoms derived from the QM9 data set, using the B3LYP functional with the 6–31G(2df,p) basis set for optimization. Low-fidelity IPs for these structures were obtained using two density functional methods: B3LYP/6–31+G(2df,p) and ωB97XD/6–311+G(3df,2p). Highly accurate G4MP2 calculations were performed on these optimized structures to obtain high-fidelity IPs to use in ML models based on the low-fidelity IPs. Our best performing ML methods gave IPs of organic molecules within a mean absolute deviation of 0.035 eV from the G4MP2 IPs for the whole data set. This work demonstrates that ML predictions assisted by quantum chemical calculations can be used to successfully predict IPs of organic molecules for use in high throughput screening.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2020

Quantified uncertainties in fission yields from machine learning

As machine learning methods gain traction in the nuclear physics community, especially those methods that aim to propagate uncertainties to unmeasured quantities, it is important to understand how the uncertainty in the training data coming either from theory or experiment propagates to the uncertainty in the predicted values. Gaussian Processes and Bayesian Neural Networks are being more and more widely used, in particular to extrapolate beyond measured data. However, studies are typically not performed on the impact of the experimental errors on these extrapolated values. In this work, we focus on understanding how uncertainties propagate from input to prediction when using machine learning methods. We use a Mixture Density Network (MDN) to incorporate experimental error into the training of the network and construct uncertainties for the associated predicted quantities. Systematically, we study the effect of the size of the experimental error, both on the reproduced training data and extrapolated predictions for fission yields of actinides.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

DOE OSTI2025

Evaluating the factors influencing accuracy, interpretability, and reproducibility in the use of machine learning classifiers in biology to enable standardization

The complexity and variability of biological data has promoted the increased use of machine learning methods to understand processes and predict outcomes. These same features complicate reliable, reproducible, interpretable, and responsible use of such methods, resulting in questionable relevance of the derived. outcomes. Here we systematically explore challenges associated with applying machine learning to predict and understand biological processes using a well- characterized in vitro experimental system. We evaluated factors that vary while applying machine learning classifers: (1) type of biochemical signature (transcripts vs. proteins), (2) data curation methods (pre- and post-processing), and (3) choice of machine learning classifier. Using accuracy, generalizability, interpretability, and reproducibility as metrics, we found that the above factors significantly mod- ulate outcomes even within a simple model system. Our results caution against the unregulated use of machine learning methods in the biological sciences, and strongly advocate the need for data standards and validation tool-kits for such studies.

59 BASIC BIOLOGICAL SCIENCES↗