Allan B. Wollaber

Massachusetts Institute of Technology | MIT · MIT Lincoln Laboratory

Recently, uncertainty-aware deep learning methods for multiclass labeling problems have been developed that provide calibrated class prediction probabilities and out-of-distribution (OOD) indicators, letting machine learning (ML) consumers and engineers gauge a model's confidence in its predictions. However, this extra neural network prediction inf...

This chapter explores military aspects of building and deploying AI Agents for cyber defense. We concentrate on those aspects which are particularly characteristic of AI Agent deployment at the “tactical edge,” by which we mean warfighters directly involved in executing the mission. This choice of emphasis is motivated by two observations. First, t...

With the increasing prevalence of encrypted network traffic, cyber security analysts have been turning to machine learning (ML) techniques to elucidate the traffic on their networks. However, ML models can become stale as new traffic emerges that is outside of the distribution of the training set. In order to reliably adapt in this dynamic environm...

This note is intended to serve as a straightforward reference that summarizes and expands on the linear aeromagnetic compensation model first introduced by Tolles and Lawson in 1950. The Tolles-Lawson model provides a simple, physical representation of an aircraft's magnetic field, composed of permanent, induced, and eddy current terms, and applies...

Through a series of federal initiatives and orders, the U.S. Government has been making a concerted effort to ensure American leadership in AI. These broad strategy documents have influenced organizations such as the United States Department of the Air Force (DAF). The DAF-MIT AI Accelerator is an initiative between the DAF and MIT to bridge the ga...

The Workshop Program of the Association for the Advancement of Artificial Intelligence’s Thirty-Sixth Conference on Artificial Intelligence was held virtually from February 22 – March 1, 2022. There were thirty-nine workshops in the program: Adversarial Machine Learning and Beyond, AI for Agriculture and Food Systems, AI for Behavior Change, AI for...

With the increasing prevalence of encrypted network traffic, cyber security analysts have been turning to machine learning (ML) techniques to elucidate the traffic on their networks. However, ML models can become stale as known traffic features can shift between networks and as new traffic emerges that is outside of the distribution of the training...

The workshop will focus on the application of AI to problems in cyber security. Cyber systems generate large volumes of data, utilizing this effectively is beyond human capabilities. Additionally, adversaries continue to develop new attacks. Hence, AI methods are required to understand and protect the cyber domain. These challenges are widely studi...

Multiple techniques for producing calibrated predictive probabilities using deep neural networks in supervised learning settings have emerged that leverage approaches to ensemble diverse solutions discovered during cyclic training or training from multiple random starting points (deep ensembles). However, only a limited amount of work has investiga...

Understanding the processes followed by organizations is important to ensure business outcomes are achieved in an optimal, efficient and compliant manner. Process mining techniques rely on the existence of structured event logs captured by process management systems. These systems are not always employed and may not capture all process steps, leavi...

Cyber network defenders face an overwhelming volume of software vulnerabilities. Resource limitations preclude them mitigating all but a small number of vulnerabilities on an enterprise network, so proper prioritization of defensive actions are of paramount importance. Current methods of risk prioritization are predominantly expert-based, and many...

Data sets that provide a ground truth to quantify the efficacy of automated algorithms are rare due to the time consuming and expensive, although highly valuable, task of manually annotating observations. These datasets exist for niche problems in developed fields such as Natural Language Processing (NLP) and Business Process Mining (BPM), however...

Consensus keyword labels. This comma-separated-value (CSV) file contains the consensus keyword labels for each of the 250 emails, indexed by local email ID. It also contains the jth Delphi member’s individual vote for keyword k under the heading “Keyword_k_member_j”. (CSV)

Consensus metalabels. This CSV file contains the consensus metalabels for each of the 250 emails, indexed by local email ID. (CSV)

Workflow construction data, CSV format. This CSV correlates the email IDs, times, trace IDs, and actions to enable the construction of workflows. It can, for instance, be used as an input file to bupaR, “Business Process Analysis in R”, available at https://www.bupar.net/. (CSV)

Workflow construction data, XES format. Upon reviewer request, the workflow construction data in 8 is also provided in the eXtensible Event Stream (XES) format defined at http://xes-standard.org. (XES)

Consensus traces. This CSV file contains the 65 consensus traces for each of the 250 emails. Each trace is indicated by a sequence of email IDs. (CSV)

Consensus workflow and action labels. This CSV file corresponds with the traces file in the S4 File (8), but the traces are assigned to workflow labels and the email IDs are replaced with consensus action labels. (CSV)

Business process models are used to identify control-flow relationships of tasks extracted from information system event logs. These event logs may fail to capture critical tasks executed outside of regular logging environments, but such latent tasks may be inferred from unstructured natural language texts. This paper highlights two workflow discov...

In this work we develop a set of nonlinear correction equations to enforce a consistent time-implicit emission temperature for the original semi-implicit IMC equations. We present two possible forms of correction equations: one results in a set of non-linear, zero-dimensional, non-negative, explicit correction equations, and the other results in a...

To assess the effectiveness of optical emission as a probe of spatial asymmetry in core-collapse supernovae (CCSNe), we apply the radiative transfer software, \supernu, to a unimodal CCSN model. The \snsph\ radiation-hydrodynamics software was used to simulate an asymmetric explosion of a 16 M$_{\odot}$ ZAMS binary star. The ejecta has 3.36 M$_{\od...

To assess the effectiveness of optical emission as a probe of spatial asymmetry in core-collapse supernovae (CCSNe), we apply the radiative transfer software, \supernu, to a unimodal CCSN model. The \snsph\ radiation-hydrodynamics software was used to simulate an asymmetric explosion of a 16 M$_{\odot}$ ZAMS binary star. The ejecta has 3.36 M$_{\od...

The Common Vulnerability Scoring System (CVSS) has been widely used to provide a score measuring the severity of software vulnerabilities. Analysts determine ordinal label assignments of subcategories relating to ease and impact of exploitation; the CVSS Version 2 (CVSS v2) score is computed formulaically using the labels. These scores have been di...

The electromagnetic transients accompanying compact binary mergers ($\gamma$-ray bursts, afterglows and 'macronovae') are crucial to pinpoint the sky location of gravitational wave sources. Macronovae are caused by the radioactivity from freshly synthesised heavy elements, e.g. from dynamic ejecta and various types of winds. We study macronova sign...

The electromagnetic transients accompanying compact binary mergers ($\gamma$-ray bursts, afterglows and 'macronovae') are crucial to pinpoint the sky location of gravitational wave sources. Macronovae are caused by the radioactivity from freshly synthesised heavy elements, e.g. from dynamic ejecta and various types of winds. We study macronova sign...

Recent efforts at Los Alamos National Laboratory to develop a moment-based, scale-bridging [or high-order (HO)-low-order (LO)] algorithm for solving large varieties of the transport (kinetic) systems have shown promising results. A part of our ongoing effort is incorporating this methodology into the framework of the Eulerian Applications Project t...

The high-order low-order (HOLO) method is a recently developed moment-based acceleration scheme for solving time-dependent thermal radiative transfer problems, and has been shown to exhibit orders of magnitude speedups over traditional time-stepping schemes. However, a linear stability analysis by Haut et al. (2015) revealed that the current formul...

The non-linear thermal radiative-transfer equations can be solved in various ways. One popular way is the Fleck and Cummings Implicit Monte Carlo (IMC) method. The IMC method was originally formulated with piecewise-constant material properties. For domains with a coarse spatial grid and large temperature gradients, an error known as numerical tele...

Downloads available at http://www.tandfonline.com/eprint/WZprCHRSEURPN7Aq3ZyR/full In 1971, Fleck and Cummings derived a system of equations to enable robust Monte Carlo simulations of time-dependent, thermal radiative transfer problems. Denoted the “Implicit Monte Carlo” (IMC) equations, their solution remains the de facto standard of high-fideli...

We present a moment-based acceleration algorithm applied to Monte Carlo simulation of thermal radiative-transfer problems. Our acceleration algorithm employs a continuum system of moments to accelerate convergence of stiff absorption–emission physics. The combination of energy-conserving tallies and the use of an asymptotic approximation in optical...

Recent efforts at Los Alamos National Laboratory to develop a moment-based, scale-bridging algorithm (or High-Order, Low-Order, HO-LO) for solving large varieties of the transport (kinetic) systems have shown promising results. A part of our ongoing effort is incorporating this methodology into the framework of the Eulerian Application Project (EAP...

50 downloads available at: http://www.tandfonline.com/eprint/qGySBT2fKsdGrp97Rpdc/full We have extended a Monte Carlo-based, moment-based acceleration algorithm to the solution of multifrequency thermal radiative transfer problems. This study focuses on two aspects. First, we consider stability/accuracy issues for a predictor-corrector time-steppin...

50 downloads available at http://www.tandfonline.com/eprint/Be3sGzs5hMIYSrYyhv8Y/full DOI: 10.1080/00411450.2014.910232 The laboratory-frame, material-motion corrections due to Morel (2006) are generalized to include a term that transports the radiation angular intensity at a multiple of the material flow speed. The generalization is shown to reta...

It is well known that temperature solutions of the Implicit Monte Carlo (IMC) equations can exceed the external boundary temperatures, a violation of the “maximum principle.” Previous attempts to prescribe a maximum value of the time-step size Δt that is sufficient to eliminate these violations have recommended a Δt that is typically too small to b...

Multigrid-preconditioned Krylov methods are applied to within-group response matrix equations of the type derived from the variational nodal method for neutron transport with interface conditions represented by orthogonal polynomials in space and spherical harmonics in angle. Since response matrix equations result in nonsymmetric coefficient matric...

A new hybrid Monte Carlo-deterministic method is presented to solve the nonlinear, frequency-dependent thermal radiative transfer (TRT) equations. Monte Carlo methods for radiative transfer typically treat temperature-dependent problem data by fixing it at the beginning of the time step, and variance reduction techniques have been limited in scope...

A new hybrid Monte Carlo-deterministic method is presented to solve the nonlinear, frequency-dependent thermal radiative transfer (TRT) equations. Monte Carlo methods for radiative transfer typically treat temperature-dependent problem data by fixing it at the beginning of the time step, and variance reduction techniques have been limited in scope...

We present physics-based preconditioning and a time-stepping strategy for a moment-based scale-bridging algorithm applied to the thermal radiative transfer equation. Our goal is to obtain (asymptotically) second-order time accurate and consistent solutions without nonlinear iterations between the high-order (HO) transport equation and the low-order...

We present an efficient numerical algorithm for solving the time-dependent grey thermal radiative transfer (TRT) equations. The algorithm utilizes the first two angular moments of the TRT equations (Quasi-diffusion (QD)) together with the material temperature equation to form a nonlinear low-order (LO) system. The LO system is solved via the Jacobi...

Implicit Monte Carlo (IMC) is a stochastic method for solving the radiative transfer equations for multiphysics application with the material in local thermodynamic equilibrium. The IMC method employs a fictitious scattering term that is computed from an implicit discretization of the material temperature equation. Unfortunately, the original histo...

Abstract not provided

It has long been known that temperature solutions of the Implicit Monte Carlo (IMC) equations can exceed the external boundary temperatures, a so-called violation of the “maximum principle.” Previous attempts at prescribing a maximum value of the time-step size ∆t that is sufficient to eliminate these violations have recommended a ∆t that is typica...

We present a new linear stability analysis of three time discretizations and Monte Carlo interpretations of the nonlinear, grey thermal radiative transfer (TRT) equations: the widely used ``Implicit Monte Carlo'' (IMC) equations, the Carter Forest (CF) equations, and the Ahrens-Larsen or ``Semi-Analog Monte Carlo'' (SMC) equations. Using a spatial...

A new hybrid Monte Carlo-deterministic method is presented to solve the nonlinear, frequency-dependent thermal radiative transfer (TRT) equations. Monte Carlo methods for radiative transfer typically treat temperature-dependent problem data by fixing it at the beginning of the time step, and variance reduction techniques have been limited in scope...

Abstract not provided

During the derivation of Fleck and Cumming's Implicit Monte Carlo (IMC) equations, a global user parameter $\alpha$ is introduced that may be adjusted in the range $0.5 \le \alpha \le 1.0$ in order to control the degree of ``implicitness'' of the IMC approximation of the thermal radiative transfer equations. For linear (and certain nonlinear) probl...

This paper presents the neutronics modeling capabilities of the fast reactor simulation system SHARP, which ANL is developing as part of the U.S. DOE's NEAMS program. We discuss the three transport solvers (PN2ND, SN2ND, and MOCFE) implemented in the UNIC code along with the multigroup cross section generation code -3. We describe the solution meth...

The UNIC code is being developed as part of the DOE's Nuclear Energy Advanced Modeling and Simulation (NEAMS) program. UNIC is an unstructured, deterministic neutron transport code that allows a highly detailed description of a nuclear reactor co re in our numerical simulations. The goal of our simulation efforts is to reduce the uncertainties a nd...

This manuscript focuses on the UNIC code, which is being developed to provide high fidelity solutions of the neutron transport equation in the fast reactor program of the Global Nuclear Energy Partnership. At present, we have developed three solvers for the neutron transport code: PN2ND, SN2ND, and MOCFE. PN2ND is based upon the second-order even-p...

A new hybrid Monte Carlo-deterministic method is presented to solve the nonlinear, frequency-dependent thermal radiative transfer (TRT) equations. Monte Carlo methods for radiative transfer typically treat temperature-dependent problem data by fixing it at the beginning of the time step, and variance reduction techniques have been limited in scope...

A new hybrid Monte Carlo-deterministic method is presented to solve the nonlinear, frequency-dependent thermal radiative transfer (TRT) equations. Monte Carlo methods for radiative transfer typically treat temperature-dependent problem data by fixing it at the beginning of the time step, and variance reduction techniques have been limited in scope...

The Method Of Characteristics (MOC) has been widely used for two dimensional lattice calculations. One of the main drawbacks of the MOC is its poor spatial representation of the within-group energy source, which requires the use of a very fine mesh to adequately resolve the neutron flux solution. The most straightforward way to improve the spatial...

A new hybrid Monte Carlo-deterministic method is presented to solve the nonlinear, frequency-dependent thermal radiative transfer (TRT) equations. Monte Carlo methods for radiative transfer typically treat temperature-dependent problem data by fixing it at the beginning of the time step, and variance reduction techniques have been limited in scope...

A quantitative theory of angular truncation errors is developed for three-dimensional discreteordinates (SN) particle transport calculations. The theory is based on an analysis of a special problem: a localized radially symmetric source in an infinite homogeneous scattering medium, with an arbitrary scattering ratio c satisfying 0 < c < 1. For both...

During the past 35 years, the Implicit Monte Carlo (IMC) method proposed by Fleck and Cummings has been the standard Monte Carlo approach to solving the thermal radiative transfer (TRT) equations. However, the IMC equations are known to have accuracy limitations that can produce unphysical solutions. In this thesis, we explicitly provide the IMC eq...

A new hybrid Monte Carlo-Deterministic technique is presented for simulating global particle transport problems, in which flux estimates are desired at all physical locations in the system. This technique has two steps: First, an inexpensive deterministic global estimate of the forward flux is obtained; then Monte Carlo is used to estimate the mult...

A new technique is presented to facilitate the Monte Carlo simulation of global particle transport problems – in which flux estimates are desired at all physical locations in the system. This technique has two steps: (i) first, an inexpensive deterministic global estimate of the forward flux is obtained, (ii) Monte Carlo is then used to estimate th...

The purpose of this research was to create computer models to expedite the core design of the International Reactor, Innovative and Secure (IRIS), specifically, so that it may employ burnable absorbers to achieve a longer cycle length and enhanced safety while minimizing the use of soluble boron. The IRIS is a next-generation, integral pressurized...