Search NASA⌕ Search

SEARCH · Search NASA

Results for “open source 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 1,567 records · Page 87

Advances in Modeling Capabilities for Critical Mineral Separation Technologies: A PrOMMiS Overview

This is an oral presentation at the TechConnect conference on the work developed by PrOMMiS. PrOMMiS builds on and extends capabilities developed within the Department of Energy’s (DOE) Institute for the Design of Advanced Energy Systems (IDAES), Integrated Platform, and Water Treatment Technoeconomic Assessment Platform (WaterTAP), which have been successfully leveraged by other Department of Energy research areas. The open-source toolkit facilitates validation, reproducibility, and accountability, allowing for easy extension of the framework to other systems. This talk presents an overview of the PrOMMiS capabilities, including unit model library, advances in thermophysical properties models, and capital cost libraries for simulation and optimization of mineral processing technologies. The PrOMMiS applications include (1) conceptual design and superstructure optimization for screening different process configurations and identifying promising technologies; (2) dynamic modeling and optimization to enable the creation of digital twins; (3) surrogate modeling tools to leverage data when predictive thermodynamic models are not currently available; (4) technical risk reduction via uncertainty quantification and robust optimization to identify process designs that are robust to process variability and uncertainties; and (5) deployment of uncertainty quantification tools to maximize knowledge gained from experimental campaigns, while reducing the number of experiments required

critical minerals and materials↗

DeePKS + ABACUS as a Bridge between Expensive Quantum Mechanical Models and Machine Learning Potentials

Recently, the development of machine learning (ML) potentials has made it possible to perform large-scale and long-time molecular simulations with the accuracy of quantum mechanical (QM) models. However, for different levels of QM methods, such as density functional theory (DFT) at the meta-GGA level and/or with exact exchange, quantum Monte Carlo, etc., generating a sufficient amount of data for training an ML potential has remained computationally challenging due to their high cost. In this work, we demonstrate that this issue can be largely alleviated with Deep Kohn–Sham (DeePKS), an ML-based DFT model. DeePKS employs a computationally efficient neural network-based functional model to construct a correction term added upon a cheap DFT model. Upon training, DeePKS offers closely matched energies and forces compared with high-level QM method, but the number of training data required is orders of magnitude less than that required for training a reliable ML potential. As such, DeePKS can serve as a bridge between expensive QM models and ML potentials: one can generate a decent amount of high-accuracy QM data to train a DeePKS model and then use the DeePKS model to label a much larger amount of configurations to train an ML potential. Further, this scheme for periodic systems is implemented in a DFT package ABACUS, which is open source and ready for use in various applications.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

System Response Testing of HEAF Kevex 130kV Microfocus Source and Thales Flashscan Panel I. Step-wedge Measurements

In the Small Sample DR Testing series, a few questions have arisen when comparing modeled results to experimental data. It is necessary to evaluate these questions in order to validate the modeling assumptions and to ensure the validity of the test procedure in the Small Sample DR Testing. Open Questions: (1) The decay time of the scintillator elements is unknown. (2) The thickness of the Beryllium window on the source is unknown. (3) The residual effects of beam angle and scattering within the detector assembly are unknown. (4) The spectral response of the scintillator and digitizing panel are not known. (5) The spectral content of the source is unknown. Item 1 will be addressed in separate tests by measuring the output of the panel at various intervals after the source has stopped illuminating the panel. Item 2 will be determined by finding the manufacturer specifications. Item 3 will be addressed in separate tests. We cannot easily address items 4 and 5 independently. Measuring the spectral responsivity of the detector directly would require the use of a calibrated reference source, which we lack. Similarly, measuring the spectral content of the source requires spectrally and spatially calibrated detectors, which we also lack. However, we can address the combination of items 1 and 2 to deduce the product of source intensity times detector response. By measuring the X-ray transmittance of samples of known thickness, material composition and density (step wedges), we can infer the desired product and so validate and improve computer models of our system.

47 OTHER INSTRUMENTATION↗

Multibody for Everybody (M4E): A Symbolic Dynamics Modeling Tool with Applications in Simulation, Control, and Optimization

Developing the analytical model of a multibody system is often the initial step in control and optimization. The analytical model (equations of motion) describes a system’s time evolution under specified forcing conditions. Although developing these equations is easy for simple systems, this process becomes more complex for systems composed of multiple bodies. Deriving equations of motion for complex multibody systems requires specialized expertise in multibody dynamics, is time-consuming, and is susceptible to error. To address this issue, this paper presents an open-source, easy-to-use, systematic framework to derive symbolic equations of motion in both Python and MATLAB using the joint coordinate formulation. This formulation results in a set of ordinary differential equations that use the minimum set of coordinates needed to model a system. The symbolic representation provides better insight into the influence of design parameters on system performance, facilitates sensitivity analysis and parameter studies, and supports direct implementation of control and optimization routines. The tool enables numerical simulation for specified parameter sets, is modular for straightforward integration with other tools and libraries, and allows incorporation of hydrodynamics, mooring, and other external forces. The result is a reproducible, extensible pipeline for modeling, simulation, and design of complex multibody systems. The proposed tool is versatile and can be applied to domains such as robotics, control, and design. In addition, we integrated external libraries that provide capabilities for modeling offshore systems such as underwater robots and marine energy converters.

16 TIDAL AND WAVE POWER↗

Open-Source and FAIR Research Software for Proteomics

Scientific discovery relies on innovative software as much as experimental methods, especially in proteomics, where computational tools are essential for mass spectrometer setup, data analysis, and interpretation. Since the introduction of SEQUEST, proteomics software has grown into a complex ecosystem of algorithms, predictive models, and workflows, but the field faces challenges, including the increasing complexity of mass spectrometry data, limited reproducibility due to proprietary software, and difficulties integrating with other omics disciplines. Closed-source, platform-specific tools exacerbate these issues by restricting innovation, creating inefficiencies, and imposing hidden costs on the community. Open-source software (OSS), aligned with the FAIR Principles (Findable, Accessible, Interoperable, Reusable), offers a solution by promoting transparency, reproducibility, and community-driven development, which fosters collaboration and continuous improvement. In this manuscript, we explore the role of OSS in computational proteomics, its alignment with FAIR principles, and its potential to address challenges related to licensing, distribution, and standardization. Drawing on lessons from other omics fields, we present a vision for a future where OSS and FAIR principles underpin a transparent, accessible, and innovative proteomics community.

97 MATHEMATICS AND COMPUTING↗

Development of a Comprehensive Physics-Based Model for Study of NASA Gateway Lunar Dust Contamination

NASA is committed to landing the first woman and first person of color on the Moon by 2025 to begin a sustaining presence. A major component of this mission is NASA’s orbiting lunar outpost, Gateway, that will be subjected to the harsh environment of space during its 7-day orbit around the Moon. One aspect of this environment that is not well quantified is microscopic lunar regolith particles, or simply, lunar dust. As Apollo 17 Mission Commander Eugene Cernan said, “I think dust is probably one of our greatest inhibitors to a nominal operation on the moon. I think we can overcome other physiological or physical or mechanical problems except dust” [1]. These dust particles, most of which are smaller than the width of two human hairs and of highly irregular shape, are composed mainly of dielectric materials. Such characteristics can allow the dust particles to collect electric charge from the surrounding environment, resulting in electrostatic interactions between the dust, electromagnetic fields, and electrically charged surfaces [2-3]. Due to the nature of these fields, dust can collect on and contaminate charged surfaces. Lunar dust introduced into the Gateway environment by a lunar ascent vehicle element returning from the surface of the Moon presents risks to Gateway hardware such as radiators, solar arrays, antennae, and docking mechanisms. To quantify this risk and inform NASA, International Partner, and Commercial Provider stakeholders, Booz Allen, in partnership with NASA, is developing a toolset to model the interaction of charged lunar dust particles with the cis-lunar and deep-space environments. This includes a comprehensive physics model of the space plasma and solar radiation environment and electromagnetic interaction of spacecraft and lunar dust. Spacecraft charging and plasma environment are solved using the open-source code from The French Aerospace Lab (ONERA) and the European Space Agency (ESA), known as Spacecraft Plasma Interaction Software (SPIS) [4]. The physics of particle charging and motion is handled by a time-dependent implementation of Orbital-Motion-Limited (OML) theory which accounts for plasma flows and positive potential grains [5-7], using Siemens STAR-CCM+ as the framework. Model validation includes lab experiments by NASA experts and academia, as well as future on-orbit dust detecting payloads on the exterior of Gateway and Commercial Lunar Payload Services (CLPS) missions to the lunar surface [7-9]. The results and analyses will inform Gateway Program system owners at risk for lunar dust contamination.

Complex Plasma↗

Methyl bromide: ocean sources, ocean sinks, and climate sensitivity

The oceans play an important role in the geochemical cycle of methyl bromide (CH3Br), the major carrier of O3-destroying bromine to the stratosphere. The quantity of CH3Br produced annually in seawater is comparable to the amount entering the atmosphere each year from natural and anthropogenic sources. The production mechanism is unknown but may be biological. Most of this CH3Br is consumed in situ by hydrolysis or reaction with chloride. The size of the fraction which escapes to the atmosphere is poorly constrained; measurements in seawater and the atmosphere have been used to justify both a large oceanic CH3Br flux to the atmosphere and a small net ocean sink. Since the consumption reactions are extremely temperature-sensitive, small temperature variations have large effects on the CH3Br concentration in seawater, and therefore on the exchange between the atmosphere and the ocean. The net CH3Br flux is also sensitive to variations in the rate of CH3Br production. We have quantified these effects using a simple steady state mass balance model. When CH3Br production rates are linearly scaled with seawater chlorophyll content, this model reproduces the latitudinal variations in marine CH3Br concentrations observed in the east Pacific Ocean by Singh et al. [1983] and by Lobert et al. [1995]. The apparent correlation of CH3Br production with primary production explains the discrepancies between the two observational studies, strengthening recent suggestions that the open ocean is a small net sink for atmospheric CH3Br, rather than a large net source. The Southern Ocean is implicated as a possible large net source of CH3Br to the atmosphere. Since our model indicates that both the direction and magnitude of CH3Br exchange between the atmosphere and ocean are extremely sensitive to temperature and marine productivity, and since the rate of CH3Br production in the oceans is comparable to the rate at which this compound is introduced to the atmosphere, even small perturbations to temperature or productivity can modify atmospheric CH3Br. Therefore atmospheric CH3Br should be sensitive to climate conditions. Our modeling indicates that climate-induced CH3Br variations can be larger than those resulting from small (+/- 25%) changes in the anthropogenic source, assuming that this source comprises less than half of all inputs. Future measurements of marine CH3Br, temperature, and primary production should be combined with such models to determine the relationship between marine biological activity and CH3Br production. Better understanding of the biological term is especially important to assess the importance of non-anthropogenic sources to stratospheric ozone loss and the sensitivity of these sources to global climate change.

NASA Discipline Exobiology↗

Lowering post‐construction yield assessment uncertainty through better wind plant power curves

Abstract Many operational analyses of wind power plants require a statistical relationship, which can be called the wind plant power curve, to be developed between wind plant energy production and concurrent atmospheric variables. Currently, a univariate linear regression at monthly resolution is the industry standard for post‐construction yield assessments. Here, we evaluate the benefits in augmenting this conventional approach by testing alternative regressions performed with multiple inputs, at a finer time resolution, and using nonlinear machine‐learning algorithms. We utilize the National Renewable Energy Laboratory's open‐source software package OpenOA to assess wind plant power curves for 10 wind plants. When a univariate generalized additive model at daily or hourly resolution is used, regression uncertainty is reduced, in absolute terms, by up to 1.0 % and 1.2 % (corresponding to a −59 % and −80 % relative change), respectively, compared to a univariate linear regression at monthly resolution; also, a more accurate assessment of the mean long‐term wind plant production is achieved. Additional input variables also reduce the regression uncertainty: when temperature is added as an input to the conventional monthly linear regression, the operational analysis uncertainty connected to regression is reduced, in absolute terms, by up to 0.5 % (−43 % relative change) for wind power plants with strong seasonal variability. Adding input variables to the machine‐learning model at daily resolution can further reduce regression uncertainty, with up to a −10 % relative change. Based on these results, we conclude that a multivariate nonlinear regression at daily or hourly resolution should be recommended for assessing wind plant power curves.

17 WIND ENERGY↗

Exploring the impact of surface topography on Rayleigh-Bénard dry convection in the Pi cloud chamber using OpenFOAM: In cylindrical and rectangular geometries

The Pi convection-cloud chamber can generate steady-state turbulence in both rectangular and cylindrical shapes via Rayleigh-Bénard convection (RBC) by maintaining warm bottom and cold top surfaces. Although most experiments in the Pi chamber were conducted in cylindrical shapes, all previous Pi chamber simulations were conducted in a rectangular shape due to the limitations of those models to discretize a cylindrical domain when using the finite difference method therein. Here, we use OpenFOAM, an open-source finite-volume-based Computational Fluid Dynamics (CFD) software package, to conduct Large-Eddy Simulation (LES) of dry RBC in the Pi chamber at high Rayleigh numbers (10 8 to 10 9 ). Results show that large-scale circulation (LSC) direction varies in the chamber with a constant side wall temperature. Imposing a slight temperature imbalance at the side wall ranging from 0.1 to 0.7 degrees can lock the LSC, aligning better with Pi chamber observations, particularly at higher Rayleigh numbers. In addition, we examine the impact of surface topography on LSC and heat transfer in RBC systems within cylindrical and rectangular shapes under varying conditions. Results show that roughing top/bottom surfaces by adding bars of a few tens millimeters height can strengthen thermal plumes and enhance temperature fluctuations in the chamber. Furthermore, we observe that different bar height configurations lead to notable changes in LSC orientation and thermal stratification, highlighting the complex interactions between surface features and convection patterns. This finding highlights how surface topography and chamber geometry affect Rayleigh-Bénard convection, improving understanding of turbulent heat transfer and atmospheric boundary-layer processes. Direct Numerical Simulations (DNS) are also conducted to validate LES results. In conclusion, while LES effectively captures qualitative behaviors seen in DNS, it tends to underestimate velocity variances near walls, illustrating a trade-off between computational efficiency and accuracy.

54 ENVIRONMENTAL SCIENCES↗

Simulation-trained machine learning models for Lorentz transmission electron microscopy

Understanding the collective behavior of complex spin textures, such as lattices of magnetic skyrmions, is of fundamental importance for exploring and controlling the emergent ordering of these spin textures and inducing phase transitions. It is also critical to understand the skyrmion–skyrmion interactions for applications such as magnetic skyrmion-enabled reservoir or neuromorphic computing. Magnetic skyrmion lattices can be studied using in situ Lorentz transmission electron microscopy (LTEM), but quantitative and statistically robust analysis of the skyrmion lattices from LTEM images can be difficult. In this work, we show that a convolutional neural network, trained on simulated data, can be applied to perform segmentation of spin textures and to extract quantitative data, such as spin texture size and location, from experimental LTEM images, which cannot be obtained manually. This includes quantitative information about skyrmion size, position, and shape, which can, in turn, be used to calculate skyrmion–skyrmion interactions and lattice ordering. We apply this approach to segmenting images of Néel skyrmion lattices so that we can accurately identify skyrmion size and deformation in both dense and sparse lattices. The model is trained using a large set of micromagnetic simulations as well as simulated LTEM images. This entirely open-source training pipeline can be applied to a wide variety of magnetic features and materials, enabling large-scale statistical studies of spin textures using LTEM.

McCray, Arthur R. C. (ORCID:0000000160774698)↗

ENZO: AN ADAPTIVE MESH REFINEMENT CODE FOR ASTROPHYSICS

This paper describes the open-source code Enzo, which uses block-structured adaptive mesh refinement to provide high spatial and temporal resolution for modeling astrophysical fluid flows. The code is Cartesian, can be run in one, two, and three dimensions, and supports a wide variety of physics including hydrodynamics, ideal and non-ideal magnetohydrodynamics, N-body dynamics (and, more broadly, self-gravity of fluids and particles), primordial gas chemistry, optically thin radiative cooling of primordial and metal-enriched plasmas (as well as some optically-thick cooling models), radiation transport, cosmological expansion, and models for star formation and feedback in a cosmological context. In addition to explaining the algorithms implemented, we present solutions for a wide range of test problems, demonstrate the code's parallel performance, and discuss the Enzo collaboration's code development methodology.

79 ASTRONOMY AND ASTROPHYSICS↗

Knowledge-guided learning with curated prior genetic biomarkers for robust model interpretation

Abstract Motivation Knowledge-guided learning offers effective and robust model training strategies in data-scarce settings by incorporating established domain knowledge, thereby enhancing generalization, robustness, and interpretability. By contrast, conventional deep learning approaches rely purely on data-driven learning, which can limit robust model interpretability, particularly in high-dimensional settings with limited size samples. In computational biology, knowledge-guided learning has primarily leveraged network- and structural-based knowledge, leading to biologically interpretable representations and enhanced predictive performance compared to conventional approaches. However, curated biomarkers, one of the most accessible forms of biological knowledge, remain largely unexplored within knowledge-guided paradigms. Results In this study, we propose a model-agnostic training paradigm, Biomarker-driven Explainable Prior-guided Learning (BioExPL), that can be applied to any neural networks that incorporates curated prior knowledge. BioExPL enforces neural networks to reflect curated biomarker priors in their latent representations through a novel knowledge-alignment loss. BioExPL consistently demonstrated significantly improved predictive performance and enhanced model interpretability with minimized computational overhead in simulation studies and intensive experiments on multiple cancer datasets. BioExPL not only integrates prior curated knowledge into the model but also accurately identifies unknown associated signals additionally. BioExPL is model-agnostic and domain-independent, enabling its integration into diverse neural network architectures. Availability and implementation The open-source is publicly available at: https://github.com/datax-lab/BioExPL.

Baek, Beomsu [Department of Computer Science, Univ↗

Visualizing effective potentials and using the IBM-Q to study quantum field theory models in 0+1 dimensions

This work explores the utility of two novel techniques applied to studying quantum field theory, namely quantum computing and visualizing effective potentials. We studied the ground state properties of the anharmonic oscillator and the Wess-Zumino model using the variational quantum eigensolver algorithm. We performed calculations for three and four qubit systems using IBM’s open source quantum computing simulation software. Using effective potentials, we study a simple dark mat-ter system with a Higgs and dark Higgs like particle interacting through a porthole fermion.

97 MATHEMATICS AND COMPUTING↗

Numerical Analysis of Detonability Assessment in a Natural Gas-Air Fueled Rotating Detonation Engine

Rotating Detonation Engines (RDE) offer an alternative combustion strategy to replace conventional constant pressure combustion with a process that could produce a pressure gain without the use of a mechanical compressor. Recent numerical and experimental publications that consider air as the oxidizer have primarily focused on the ability of these annular combustors to sustain a stable continuous detonation wave when fueled by hydrogen. However, for this to be a viable consideration for the land-based power generation it is necessary to explore the ability to detonate natural gas and air within the confines of the annular geometry of an RDE. Previous studies on confined detonations have expressed the importance of permitting detonation cells to fully form within the combustor in order to achieve stability. This poses a challenge for natural gas–air fueled processes as their detonation cell size can be quite large even at moderate pressures. Despite the practical importance, only a few studies are available on natural gas detonations for air-breathing RDE applications. Moreover, the extreme thermodynamic condition (high temperature inside the combustor) allows limited accessibility inside the combustor for detailed experimental instrumentations, providing mostly single-point data. Recent experimental studies at the National Energy Technology Laboratory (NETL) have reported detonation failure at higher methane concentration in an air-breathing RDE fueled by natural gas-hydrogen fuel blends. This encourages to perform a detailed numerical investigation on the wave characteristics of detonation in a natural gas-air fueled RDE to understand the various aspects of instability associated with the natural gas-air detonation.This study is a numerical consideration of a methane-air fueled RDE with varying operating conditions to ascertain the ability to achieve a stable, continuous detonation wave. The simulations have been performed in a 2D unwrapped RDE geometry using the open-source CFD library “OpenFOAM” employing an unsteady pressure-based compressible reactive flow solver with a <em>k</em>–<em>ε</em> turbulence model in a structured rectangular grid system. Both reduced and detailed chemical kinetic models have been used to assess the effect of the chemistry on the detonation wave characteristics and the underlying flow features. A systematic grid sensitivity study has been conducted with various grid sizes to quantify the weakly stable overdriven detonation on a coarse mesh and oscillating features at fine mesh resolutions.The main focus of the current study is to investigate the effects of operating injection pressure on detonation wave characteristics of an air-breathing Rotating Detonation Engine (RDE) fueled with natural gas-hydrogen fuel blends. Wave speeds, peak pressures and temperatures, and dominant frequencies have been computed from the time histories. The flow structures were then visualized using 2D contours of temperature and species concentration.

Saha, Pankaj↗

muMatScale

Code for modeling solidification microstructure based on the Cellular Automata method for binary/pseudo-binary alloys. This code is derived from the open-source software uMatIC developed at the Imperial College London https://www.imperial.ac.uk/engineering-alloys/research/software

Yuan, Lang [Univ. of South Carolina, Columbia, SC ↗

sup3ruhi (Super Resolution for Renewable Resource Data and Urban Heat Islands) [SWR-25-05]

Urban heat is a growing concern, particularly in dense metropolitan areas where high temperatures increase the risk of heat-related illness and drive energy expenses for cooling. Estimating the effects of urban heat remains a challenge due to limitations in describing the built environment, computational constraints, and the need for high-resolution data. This software presents open-source, computationally efficient machine learning methods that enhance the accuracy of urban temperature estimates compared to historical reanalysis data. Models trained using this software have been applied to urban microclimates in Los Angeles and Seattle showing greater accuracy and less bias when compared to low-resolution reanalysis datasets like ERA5 and even when compared to high-resolution mesoscale numerical weather models like WRF with an urban canopy model. Initial findings highlight how machine learning can support urban heat resilience planning by enabling improved assessments of local heat islands, mitigation strategies, and their energy implications. This software is an extension of (sup3r). This software supports the following publication: Buster, Grant, et al. Tackling Extreme Urban Heat: A Machine Learning Approach to Assess the Impacts of Climate Change and the Efficacy of Climate Adaptation Strategies in Urban Microclimates. arXiv:2411.05952, arXiv, 8 Nov. 2024. arXiv.org, https://doi.org/10.48550/arXiv.2411.05952. And has related public data records available at: Buster, Grant, Cox, Jordan, Benton, Brandon, and King, Ryan. Super-Resolution for Renewable Resource Data and Urban Heat Islands (Sup3rUHI). United States: N.p., 16 Oct, 2024. Web. https://data.openei.org/submissions/6220.

Buster, Grant [National Renewable Energy Laborator↗

A tutorial review of machine learning-based model predictive control methods

Abstract This tutorial review provides a comprehensive overview of machine learning (ML)-based model predictive control (MPC) methods, covering both theoretical and practical aspects. It provides a theoretical analysis of closed-loop stability based on the generalization error of ML models and addresses practical challenges such as data scarcity, data quality, the curse of dimensionality, model uncertainty, computational efficiency, and safety from both modeling and control perspectives. The application of these methods is demonstrated using a nonlinear chemical process example, with open-source code available on GitHub. The paper concludes with a discussion on future research directions in ML-based MPC.

Wu, Zhe [Department of Chemical and Biomolecular E↗

Deep PDF parsing to extract features for detecting embedded malware.

The number of PDF files with embedded malicious code has risen significantly in the past few years. This is due to the portability of the file format, the ways Adobe Reader recovers from corrupt PDF files, the addition of many multimedia and scripting extensions to the file format, and many format properties the malware author may use to disguise the presence of malware. Current research focuses on executable, MS Office, and HTML formats. In this paper, several features and properties of PDF Files are identified. Features are extracted using an instrumented open source PDF viewer. The feature descriptions of benign and malicious PDFs can be used to construct a machine learning model for detecting possible malware in future PDF files. The detection rate of PDF malware by current antivirus software is very low. A PDF file is easy to edit and manipulate because it is a text format, providing a low barrier to malware authors. Analyzing PDF files for malware is nonetheless difficult because of (a) the complexity of the formatting language, (b) the parsing idiosyncrasies in Adobe Reader, and (c) undocumented correction techniques employed in Adobe Reader. In May 2011, Esparza demonstrated that PDF malware could be hidden from 42 of 43 antivirus packages by combining multiple obfuscation techniques [4]. One reason current antivirus software fails is the ease of varying byte sequences in PDF malware, thereby rendering conventional signature-based virus detection useless. The compression and encryption functions produce sequences of bytes that are each functions of multiple input bytes. As a result, padding the malware payload with some whitespace before compression/encryption can change many of the bytes in the final payload. In this study we analyzed a corpus of 2591 benign and 87 malicious PDF files. While this corpus is admittedly small, it allowed us to test a system for collecting indicators of embedded PDF malware. We will call these indicators features throughout the rest of this report. The features are extracted using an instrumented PDF viewer, and are the inputs to a prediction model that scores the likelihood of a PDF file containing malware. The prediction model is constructed from a sample of labeled data by a machine learning algorithm (specifically, decision tree ensemble learning). Preliminary experiments show that the model is able to detect half of the PDF malware in the corpus with zero false alarms. We conclude the report with suggestions for extending this work to detect a greater variety of PDF malware.

99 GENERAL AND MISCELLANEOUS↗