AI ML | Kalyan Perumalla

CYVET

Sat, 01 Jan 2022 00:00:00 +0000

Our new Cyber-Physical Security Assurance Framework based on Semi-Supervised Vetting applies the latest AI/ML and NLP technologies on hardware testbeds to advance the resilience of critical cyber-physical assets including electric grids and gas pipelines.

CYVET Pipeline

Overview

Organization

Sponsor: US Department of Energy (DOE)
- Office: Cybersecurity, Energy Security, and Emergency Response (CESER)
- Program: Cybersecurity for Energy Delivery Systems (CEDS)
- Award: CESER
Prime: Oak Ridge National Laboratory (ORNL)
- Subcontract: University of Nebraska-Lincoln (UNL)
Period: 2019-2023

Gallery

OSTI.gov: https://www.osti.gov/servlets/purl/1661247

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

We introduce CyBERT, a cybersecurity feature claims classifier based on bidirectional encoder representations from transformers and a key component in our semi-automated cybersecurity vetting for industrial control systems (ICS)…The results showed that CyBERT outperforms these models on the validation accuracy and the F1 score, validating CyBERT’s robustness and accuracy as a cybersecurity feature claims classifier.

Kimia Ameri, Michael Hempel, Hamid Sharif, Juan Lopez, Kalyan Perumalla

PDF Cite DOI

Trust-but-Verify in Cyber-Physical Systems

Cyber-physical systems span a wide spectrum, from long-lived legacy systems to more modern installations. Trust is an issue that arises across the spectrum, albeit with different variants of goals and constraints. On the one end of the spectrum, legacy systems are characterized by function-based designs in which trust is an implicitly in-built concept…

Kalyan Perumalla

PDF Cite Slides DOI

Smart Semi-Supervised Accumulation of Large Repositories for Industrial Control Systems Device Information

A solution is needed for vetting the vendor-supplied feature claims and their adherence to cybersecurity requirements and standards. We are presently engaged in an effort to develop such a system. This paper demonstrates one vital aspect of this effort in proposing an end-to-end framework to accumulate a large repository of ICS device information for this vetting system, curate the dataset, and conduct extensive processing. This framework is designed to use web scraping, data analytics and Natural Language Processing (NLP) techniques to identify vendor websites, automate the collection of website-accessible documents and automatically derive metadata from them for identification of product documents relevant to the repository…

Kimia Ameri, Michael Hempel, Hamid Sharif, Juan Lopez, Kalyan Perumalla

PDF

A Novel Vetting Approach to Cybersecurity Verification in Energy Grid Systems

The cybersecurity auditing for Operation Technology is critical and has been largely missing from the cybersecurity research, especially in the energy sector. In this paper, we present a novel “cybersecurity vetting” approach (CYVET) to the problem of verification and validation of cybersecurity in complex cyber-physical installations underlying modern energy grid systems.

Kalyan Perumalla, Juan Lopez, Maksudul Alam, Olivera Kotevska, Michael Hempel, Hamid Sharif

PDF DOI

Additional Background

The cybersecurity auditing for Operation Technology (OT) is critical and has been largely missing from the cybersecurity research, especially in the energy sector. CYVET is a novel “cybersecurity vetting” approach (CYVET) to the problem of verification and validation of cybersecurity in complex cyber-physical installations underlying modern energy grid systems.

In Information Technology (IT), cybersecurity auditing is a widespread practice to ensure privacy, security, and trust. However, for the field of Operation Technology (OT) as used in electric energy systems, this is a relatively novel concept. In fact, OT itself only recently began to embrace IT principles, with the push for automation and centralized control driving this development. OT operators are simply not yet used to the idea of cybersecurity. To ameliorate the gap, product vendors for field devices are advancing the field by incorporating more and more security features into their products. However, customers are often either unaware of them, or do not use them, or cannot use them because of unsatisfied device ecosystem dependencies. There is thus a disconnect between what is offered, what is possible post-deployment, and what the customer expects.

There is a vast lack of cybersecurity oversight and insight, from a certification and a customer perspective alike, for OT systems in the energy sector. With new features constantly being added to new and existing products, customers are predominantly unaware what their purchased solutions are capable of, or not capable of. They often do not know if their current systems meet their own cybersecurity requirements as well as industry standards. Many of these facets not only indirectly depend on device capabilities, but also on device deployment decisions – Does a newly added feature work in an existing context? Can it be used as envisioned? Does it interfere with other cybersecurity requirements? Does it produce side effects that may interfere with other requirements?

Hence, what is needed is a security vetting system designed to provide insight into deployed systems, the match of capabilities to requirements, adherence to certification requirements, and so forth. There are few systems currently available that provide these energy grid security capabilities. OT systems are increasingly cyber-enabled, increasingly complex, and increasingly interdependent. This rapidly accelerating trend poses a clear risk for asset owners to lose confidence and for cybersecurity risk to go undiscovered until exploited by malicious parties.

Deep CYBERIA

Sat, 01 Jan 2022 00:00:00 +0000

Deep CYBERIA is a novel system focused on Detecting Sensors Deeply Embedded in Cyber-Physical Systems via novel machine learning and passive/active/hybrid probing techniques on complex operational technology (OT) networks.

Overview

Deep CYBERIA is designed to address the critical capability gap identified by USAF in discovering, identifying and mapping the edge devices and physical sensors connected to those edge devices. The effort is aimed at the outcome of providing situational awareness of Industrial Control Systems/Supervisory Control and Data Acquisition (ICS/SCADA) traffic and devices operating in a network. The situational awareness will support the requirement of performing deep dive filtering analysis and enumeration of ICS traffic.

Demonstration

Organization

Sponsor: US Department of Defense
Prime: Oak Ridge National Laboratory (ORNL)
- Subcontract: MIT Lincoln Laboratory (MIT-LL)
Period: 2019-2025

Gallery

Trust-but-Verify in Cyber-Physical Systems

Kalyan Perumalla

PDF Cite Slides DOI

Detecting Sensors and Inferring their Relations at Level-0 in Industrial Cyber-Physical Systems

In this paper, we present our research and development efforts aimed at addressing the gap in discovering sensors at level 0 in industrial CPS by building a system called Deep-cyberia (Deep Cyber-Physical System Interrogation and Analysis) that incorporates algorithms and interfaces aimed at uncovering sensors and computing estimates of correlations among them.

Kalyan Perumalla, Srikanth Yoginath, Juan Lopez

PDF Cite Slides DOI

Volatile Memory Extraction-Based Approach for Level 0-1 CPS Forensics

Our focus is to extract volatile and dynamically changing internal information form CPS 0-1 level devices, and design preliminary schemes to exploit that extracted information. As a case study, we apply the proposed methodology to Modicon PLC using Modbus protocol. We extract the memory layout and subject the device to read operations at the most critical regions of memory. This capability of generating a sequence of volatile memory snapshots for offline, detailed and sophisticated analysis opens a new class of cyber security schemes for CPS forensic analysis, taint analysis and watermarking.

Rima Asmar Awad, Juan Lopez, Michael Rogers, Kalyan Perumalla

PDF Cite

ZeroIn

Sat, 01 Jan 2022 00:00:00 +0000

Our formulation of a new AI/ML-based, real-time computational framework aims to learn and flag defects in software as early as the time of commit in the developers’ repositories.

ZeroIn Network

Overview

Using novel AI/ML techniques, ZeroIn zeroes-in onto problems in software repositories and aims to identify code vulnerabilities at their very origin, namely, at the time at which developers commit their codes into their repositories.

ZeroIn: Characterizing the Data Distributions of Commits in Software Repositories

A characterization of the software development metadata is presented in terms of distributions of data that best captures the trends in the datasets, to feed into the machine learning components of ZeroIn to exploit connectivity among the sets of repositories, commits, and developers.

Kalyan Perumalla, Aradhana Soni, Rupam Dey, Steven Rich

PDF

Using Machine Learning Towards Early Flagging of Potentially Buggy Software Commits

Using multiple classifiers we verify the feasibility of using metadata from synthetic datasets modeled by a characterization of a few large software repositories and developer profiles. Results show that the metadata-based learning approach appears promising towards early flagging of potentially buggy commits in software repositories.

Aradhana Soni, Kalyan Perumalla

PDF

Characterizing the Distributions of Commits in Large Source Code Repositories

We present preliminary results from characterizing the distribution of 452 million commits in a metadata listing from GitHub repositories. Based on multiple distributions, we find the best fits and second best fits across different ranges in the data. The characterization is aimed at synthetic repository generation suitable for use in simulation and machine learning.

Aradhana Soni, Kalyan Perumalla, Rupam Dey

PDF

Organization

Sponsor: Industry
Institutions: University of Tennessee, Knoxville
Period: 2021-2024

Gallery

AI ML Research

Wed, 01 Dec 2021 00:00:00 +0000

My research interests in AI/ML have focused on applying the technologies and software to applications in cyber-physical systems, digital twins, cybersecurity verification, and intelligent mechanical design. My research has largely used existing AI/ML methods and using them effectively in novel ways to solve difficult problems in various domains. Nevertheless, I am interested in the fundamental concepts underlying the methods and developing computational implementations from scratch, as, for example, in new Recurrent Neural Network implementations from scratch, not entirely relying on popular AI/ML packages.

From the computational point of view, my other major interests are in the end-to-end scalability of AI/ML solutions, especially in relation to novel hardware architectures, down-to-the-metal considerations, hardware-software interplay and codesign, and holistic scalability of workflows. Of particular interest are new AI/ML problems that are characterized by volumes and velocities massively higher than what typical desktop-based and cluster-based solutions address.

While reaping the benefits of well known and tested methods, my inclination is to keep an eye towards basic insights, unique combinations, and new techniques.

Overview

Projects

Deep CYBERIA
- CNN: Sensor detection
- RNN: Sensor time-series
CYVET
- NLP: CyBERT over BERT
- NLP: Claim detection
- CNN: Document classification
- CNN: Product detection
Digital Twin Framework
- RNN: Anomaly detection
DeepEx
- CNN: Material image classification
- VAE: Molecular Dynamics
Intelligent Design
- RL: Steering
Energy Management
- RL: Dynamic learning and adaptation
Kensor
- Critical Facility Monitoring
- Novel Markov modeling

Software

Python
Jupyter
numpy/no-numpy
CUDA
CUDNN
MPI
Tensorflow (2/Keras)
PyTorch

Organization

Sponsors: Various, across multiple projects
- US Department of Energy (DOE)
- US Department of Defense (DoD)
- Oak Ridge National Laboratory-Directed Research and Development (LDRD)
- Industry

Publications

Listed under the AI/ML category under Publications At-a-glance

Digital Twin Framework

Sat, 01 Jan 2022 00:00:00 +0000

Our novel Digital Twin Framework (DTF) is designed to improve resilience of critical infrastructure systems by continuously comparing the infrastructure state with automatically generated, AI/ML-based, real-time digital-twin simulation of the system.

Team

Digital Twin Framework Software Architecture

Overview

As the level of automation in critical infrastructure increases, the ability to detect cyber intrusions becomes more crucial and extremely challenging. Recent cyber attacks demonstrate the devastating and widespread affects they can have on critical infrastructure.

DTF is designed specifically to detect and eventually prevent such attacks, with models validated against experimental data from two critical infrastructure experimental emulators – a canal lock system and an electric distribution system – exhibiting very different dynamics. The canal lock system’s digital twin uses a recurrent neural network trained from the experimental data collected via the DTF. A digital twin of the transmission system is created using a commercial real-time power systems simulator and integrated into our DTF along with the hardware, embedded controllers, and live sensor data using the Open Field Message Bus data model, and publish/subscribe communication protocols. A cyber attack is used on both systems to demonstrate the DTF’s detection capability.

Organization

Sponsor: Lab Directed Research and Development, Oak Ridge National Laboratory
Period: 2018-2020

Gallery

On the Effectiveness of Recurrent Neural Networks for Live Modeling of Cyber-Physical Systems

We empirically study the effectiveness of Recurrent Neural Network (RNN)-based models as the basis of DT-based resilience and uncover the important characteristics of an RNN-based solution with experimentation on a lab-scale Canal Lock CPS emulator with live validations and attack scenarios. For the first time, we demonstrate actual, real-time use of a RNN-based model as a DT for performing live analysis on an operational CPS.

Srikanth Yoginath, Varisara Tansakul, Supriya Chinthavali, Curtis Taylor, Joshua Hambrick, Philip Irminger, Kalyan Perumalla

PDF Cite DOI

A Digital Twin Framework for Testing, Evaluation and Deployment of Resilient Cyber-physical Systems

We describe an approach to detecting and preventing cyber attacks by continuously comparing the infrastructure state with a real-time digital-twin simulation of it. Specifically, we describe and demonstrate a Digital Twin Framework (DTF) designed specifically to detect and eventually prevent such attacks. The canal lock system’s digital twin uses a recurrent neural network trained from the experimental data collected via the DTF.

Raymond Hink, Mark Buckner, Supriya Chinthavali, Chris Craig, Timothy Daniel, Joel Dawson, Milton Ericson, Joshua Hambrick, Philip Irminger, Ryan Kerekes, Juan Lopez, Kalyan Perumalla, Nicholas Peters, Stacy Prowell, Varisara Tansakul, Curtis Taylor, Bailu Xiao, Srikanth Yoginath

PDF

Kensor: Coordinated Intelligence from Co-located Sensors

Sat, 01 Jan 2022 00:00:00 +0000

Kensor uncovers coordinated intelligence about normal and abnormal phenomena from multiple sensors co-located in close proximity in secure installations.

Kensor

Overview

Given a set of co-located sensors, we seek an intelligent approach that would automatically determine the “normal” patterns of behaviors among the correlated sensors.

After normal behavior is extracted, later monitoring should detect any deviant variations over time.

An example application is an entry monitoring and alert system for facilities such as nuclear reactors, where badge readers, door locks, lights, weight trackers and other co-located sensors at the entry point are collectively tracked.

Gallery

Organization

Sponsor: Nuclear Safety
Period: 2019-2020

Kensor: Coordinated Intelligence from Co-Located Sensors

Here, we focus on coordinated intelligence about normal and abnormal phenomena from multiple sensors geographically co-located, monitoring and controlling a set of co-located devices. Given a set of co-located sensors, we develop an intelligent approach that automatically determines the ’normal’ patterns of behaviors among the correlated sensors. After normal behavior is extracted, later monitoring detects deviant variations over time.

Olivera Kotevska, Kalyan Perumalla, Juan Lopez

PDF Cite DOI

Advancements in Artificial Intelligence for Science

Thu, 01 Feb 2024 00:00:00 +0000

Solicitation PDF

Original at OSTI
Cached local copy

Selected Pages

Selected Extracts

The DOE SC program in Advanced Scientific Computing Research (ASCR) hereby announces its interest in basic computer science and applied mathematics research in the fundamentals of Artificial Intelligence (AI) for science. Specifically, advancements in this area are sought that can enable the development of:

Foundation models for computational science;
Automated scientific workflows and laboratories;
Scientific programming and scientific-knowledge-management systems;
Federated and privacy-preserving training for foundation and other AI models for science; and
Energy-efficient AI algorithms and hardware for science.

The development of new AI techniques applicable to multiple scientific domains can accelerate progress, increase transparency, and open new areas of exploration across the scientific enterprise.

Research Area 1: …

…

Research Area 2: AI Innovations for Scientific Knowledge Synthesis and Software Development

The state-of-the-art in knowledge synthesis and programming tools are changing rapidly, fueled by AI Large Language Models (LLMs) trained on text, source code, and other data sources. New AI-driven tools are currently not trustworthy; do not systematically understand mathematical and physical principles; cannot properly ingest and understand scientific literature and data; and do not produce consistent, verified, uncertainty-quantified, reproducible results. In addition to addressing those challenges, there may be particular advantages in such tools building up knowledge and context over many interactions with a user or group of users. However, incremental training of AI systems over long time horizons, and the representation of knowledge in AI systems robust to changes in the underlying AI models, remain critical challenges. This research area seeks fundamental advancements in knowledge synthesis and programming tools for science. Moreover, realizing AI systems that can truly understand, and assist with, all aspects of the scientific process requires innovation in many areas, including multimodality, tool use, deeper reasoning and planning, memory, and external interaction. For additional background, see Chapter 2, “AI Foundation Models for Scientific Knowledge Discovery, Integration, and Synthesis,” Chapter 6, “AI for Programming and Software Engineering,” Chapter 12, “Mathematics and Foundations,” and Chapter 14, “Data Ecosystem,” of the AI For Science, Energy, and Security report [1].

Additionally, investigations into AI-driven tools for science should be conceptualized accounting for the iterative and collaborative processes that define modern science and scientific-software development. Accordingly, research proposed in this area is encouraged to address the relevant Priority Research Directions (PRDs) from the Basic Research Needs in The Science of Scientific Software Development and Use report [6], which are PRD 1, “Develop next-generation tools to enhance developer productivity and software sustainability,” PRD 2, “Develop methodologies and tools to comprehensively improve team-based scientific software development and use,” and PRD 3, “Develop methodologies, tools, and infrastructure for trustworthy software-intensive science.

Methods proposed for investigation should use any appropriate techniques that might be necessary to accomplish their goals, including, but not limited to, machine learning, natural- language processing, formal reasoning, instrumentation, data management, and compiler technology. The sustainability and explainability of scientific software are critically important to the scientific process, and as a result, particular consideration should be given to maximizing the extent to which human programmers understand and/or trust the outputs of these methods.

Research Area 3: …

…

Research Area 4: …

…

Research Area 5: …

…

Artificial Intelligence Tools for Catalyzing Interdisciplinary Science (SBIR)

Sat, 01 Jul 2023 00:00:00 +0000

Solicitation PDF

Original at OSTI
Cached local copy

Selected Pages

Selected Extracts

This topic is focused on new artificial intelligence (AI) tools that enhance the productivity of scientists and engineers when making use of scholarly publications and engaging in interdisciplinary interactions in the areas of science and engineering supported by SC. While scientists are deeply knowledgeable in their areas of expertise, they are often not as highly informed in other important disciplines. For example, experts in several subareas of high-performance computing are not as deeply knowledgeable about the terminology, concepts, and state of the art in other areas such as biological sciences or high energy physics. The advent of new service-oriented access to technology that is based on large language models (LLM), which may include multi- modal data, has now opened the possibility of utilizing LLM-based commercial services to enable new interdisciplinary synergy. These services can now be envisioned to be used to digest large amounts of scientific publications and documentation across disciplines and enable interdisciplinary interactions that were not conceivable before.

Against this backdrop, grant applications are sought on the topic of “AI Tools for Catalyzing Interdisciplinary Science.” This topic will be focused on increasing the synergy among disciplines supported by the Office of Science. For example, this would include AI-based tools that can catalyze the interactions among scientists in nuclear physics and material sciences. Innovative methods are needed to assimilate the scientific publication corpus of two or more disciplines and enable scientists in any of those disciplines to collate scientific ideas, concepts, questions, and solutions from the other disciplines.

Included in scope is the integration with commercial AI services (Google, Microsoft, OpenAI, etc.) and open/commercial sources of publications and other data. Proposed approaches must display a short-term path to success and commercial viability. The proposed work should include a plan to perform demonstration activities with scientists and demonstrate verification and validation of results, including data validation.

Grant applications focused on the following will be considered out of scope:

Tools that address less than two scientific disciplines in Office of Science research areas.
Tools that build LLM-based services from scratch.
Security and hardening of LLM.

a. Interdisciplinary Training and Interfaces

Applications responsive to this subtopic will address the challenge of ingesting a multi-modal corpus of scientific publications of two or more scientific disciplines in Office of Science research areas into a knowledgebase to be built using LLM-based services.

Additionally, applications may address the creation of modern interfaces with natural language-based prompt- and-response support necessary for scientists to interact with the LLM back-ends of interdisciplinary knowledgebases.

Determining the Most Significant Metadata Features to Indicate Defective Software Commits

Tue, 23 May 2023 00:00:00 +0000

Impact of Grammar on Language Model Comprehension

Mon, 20 Feb 2023 00:00:00 +0000

https://www.computer.org/csdl/proceedings-article/icnc/2023/10074239/1LKwLebVPGw

Design of a Novel Information System for Semi-Automated Management of Cybersecurity in Industrial Control Systems

Wed, 22 Jun 2022 00:00:00 +0000

Article is available as Open Access

An Accuracy-Maximization Approach for Claims Classifiers in Document Content Analytics for Cybersecurity

Wed, 15 Jun 2022 00:00:00 +0000

Open Access: https://www.mdpi.com/2624-800X/1/4/31.

ZeroIn: Characterizing the Data Distributions of Commits in Software Repositories

Sat, 16 Apr 2022 00:00:00 +0000

Using Machine Learning Towards Early Flagging of Potentially Buggy Software Commits

Fri, 08 Apr 2022 00:00:00 +0000

Characterizing the Distributions of Commits in Large Source Code Repositories

Wed, 15 Dec 2021 00:00:00 +0000

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

Thu, 04 Nov 2021 00:00:00 +0000

Open Access: https://www.mdpi.com/2624-800X/1/4/31.
This article belongs to the Special Issue Machine Learning and Data Analytics for Cyber Security.

Trust-but-Verify in Cyber-Physical Systems

Wed, 28 Apr 2021 00:00:00 +0000

https://dl.acm.org/doi/abs/10.1145/3445969.3450434

Smart Semi-Supervised Accumulation of Large Repositories for Industrial Control Systems Device Information

Mon, 01 Feb 2021 00:00:00 +0000

A Novel Vetting Approach to Cybersecurity Verification in Energy Grid Systems

Mon, 13 Jul 2020 00:00:00 +0000

https://www.osti.gov/servlets/purl/1661247

https://ieeexplore.ieee.org/abstract/document/9167562

COVID-relevant Scalable Computational Research Directions and Tools

Thu, 09 Apr 2020 00:00:00 +0000

This seminar conveys some directions and technical ideas being pursued by the Discrete Computing Systems Group of the Computer Science and Mathematics Division, in collaboration with others at the lab. Some of the scalable computational tools ready for applying to challenging computational problems are presented. Actual working codes ready for customization to COVID-related efforts are described, which are built for scaling to supercomputing platforms such as Summit.

Topics that are covered include:

A high resolution simulator, ExaCorona, that scales from laptops to leadership class supercomputers, is outlined that uses a discrete event model of virus spread, with probabilistically timed state transitions at the individual level across millions of individuals represented with arbitray geography and mobility characteristics.
A clonable simulation framework, CloneX, is introduced that enables millions of “what-if” scenarios to be executed rapidly on thousands of GPUs of Summit and similar supercomputers. An SEIR-based epidemiological model is outlined for numerous what-if simulations of disease spread that can be executed for country-scale populations like India’s, with ‘what-if’ scenarios, each varying in the outbreak points (hotspots), quarantines, vaccinations and hospitalizations.
A machine learning pipeline for the prediction of material structure properties directly from their neutron scattering profiles (development as part of the ExaLearn ECP co-design project). A brief overview of this system is provided, which is being applied for studies of new therapeutic targets and viral protein-structure-assisted drug design studies related to the COVID outbreak.
Network science methods are outlined for detecting information cascades in time varying large-scale social communication networks. We discuss its implications for detecting occurrence/response or epidemic related events from Twitter and similar global interaction systems.

RL-HEMS: Reinforcement Learning-Based Home Energy Management System for HVAC Energy Optimization (OR-20-C051)

Tue, 04 Feb 2020 00:00:00 +0000

https://ashraem.confex.com/ashraem/w20/meetingapp.cgi/Paper/26352

A Digital Twin Framework for Testing, Evaluation and Deployment of Resilient Cyber-physical Systems

Wed, 01 Jan 2020 00:00:00 +0000

RL-HEMS: Reinforcement Learning Based Home Energy Management System for HVAC Energy Optimization.

Wed, 01 Jan 2020 00:00:00 +0000

Kensor: Coordinated Intelligence from Co-Located Sensors

Mon, 09 Dec 2019 00:00:00 +0000

https://ieeexplore.ieee.org/abstract/document/9006318

On the Effectiveness of Recurrent Neural Networks for Live Modeling of Cyber-Physical Systems

Mon, 11 Nov 2019 00:00:00 +0000

https://ieeexplore.ieee.org/abstract/document/9065023

Detecting Sensors and Inferring their Relations at Level-0 in Industrial Cyber-Physical Systems

Tue, 05 Nov 2019 00:00:00 +0000

https://ieeexplore.ieee.org/abstract/document/9032891

Towards Native Execution of Deep Learning on a Leadership-Class HPC System

Mon, 20 May 2019 00:00:00 +0000

https://ieeexplore.ieee.org/abstract/document/8778212

AI ML | Kalyan Perumalla

CYVET

Overview

Organization

Gallery

Related Publications

Additional Background

Deep CYBERIA

Overview

Demonstration

Organization

Gallery

Related Publications

ZeroIn

Overview

Related Publications

Organization

Gallery

AI ML Research

Overview

Projects

Software

Organization

Publications

Digital Twin Framework

Overview

Organization

Gallery

Related Publications

Kensor: Coordinated Intelligence from Co-located Sensors

Overview

Gallery

Organization

Related Publications

Advancements in Artificial Intelligence for Science

Solicitation PDF

Selected Pages

Selected Extracts

Artificial Intelligence Tools for Catalyzing Interdisciplinary Science (SBIR)

Solicitation PDF

Selected Pages

Selected Extracts

a. Interdisciplinary Training and Interfaces

Determining the Most Significant Metadata Features to Indicate Defective Software Commits

Impact of Grammar on Language Model Comprehension

Design of a Novel Information System for Semi-Automated Management of Cybersecurity in Industrial Control Systems

An Accuracy-Maximization Approach for Claims Classifiers in Document Content Analytics for Cybersecurity

ZeroIn: Characterizing the Data Distributions of Commits in Software Repositories

Using Machine Learning Towards Early Flagging of Potentially Buggy Software Commits

Characterizing the Distributions of Commits in Large Source Code Repositories

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

Trust-but-Verify in Cyber-Physical Systems

Smart Semi-Supervised Accumulation of Large Repositories for Industrial Control Systems Device Information

A Novel Vetting Approach to Cybersecurity Verification in Energy Grid Systems

COVID-relevant Scalable Computational Research Directions and Tools

RL-HEMS: Reinforcement Learning-Based Home Energy Management System for HVAC Energy Optimization (OR-20-C051)

A Digital Twin Framework for Testing, Evaluation and Deployment of Resilient Cyber-physical Systems

RL-HEMS: Reinforcement Learning Based Home Energy Management System for HVAC Energy Optimization.

Kensor: Coordinated Intelligence from Co-Located Sensors

On the Effectiveness of Recurrent Neural Networks for Live Modeling of Cyber-Physical Systems

Detecting Sensors and Inferring their Relations at Level-0 in Industrial Cyber-Physical Systems

Towards Native Execution of Deep Learning on a Leadership-Class HPC System