Deep Models and Artificial Intelligence for Military
Applications: Potentials, Theories, Practices, Tools and Risks
With advancements in computer storage capacity and parallel
processing, Big Data has become omnipresent. Related to big data is Deep
Analytics, which includes machine learning (ML), deep learning (DL) and
artificial intelligence (AI). These methods and tools are abundant in the
commercial world. However, they may not be appropriate to solve many military
problems. Military applications require data sources that are distributed,
disparate, multi-sourced and real-time and are of extremely high rates, high
volumes and high varieties. The needs for information sharing and agility as
well as strict security, likely multi-layers with guarantees of effect in the
case of intrusion, as well as redundancy to severed resources, across all
domains make the problem more complex. These problems often require new
mathematical approaches to optimization. This symposium explores the
potential benefits of ML/DL/AI methods and systems to military applications.
Traditional data sciences including statistics, numerical analysis, machine
learning, data mining, business intelligence, and artificial intelligence have evolved
into Big Data analytics. The current technologies are dominated by
systems that provide 1) Secure and distributed data storage; 2) parallel and
distributed computing; and 3) Deep Analytics.
Current Deep Analytics essentially include same or similar algorithms and methodologies in the traditional statistics, numerical analysis, machine learning, data mining, business intelligence, and artificial intelligence, however, they are scaled up to Big Data. As the data size gets bigger, the statistical significance of the analytics is often guaranteed due purely to the size. This positive impact of the data size can be a great advantage. However, other challenges rise. For example, traditional data sciences used in small- or moderate-sized, analysis typically require tight coupling of the computations. Such an algorithm often executes in a single machine or job and reads all the data at once. How can these algorithms be modified so they can be executed in parallel in thousands of clusters? These challenges in theories and in practices are not only for military applications, but also the challenges for all Big Data applications.
For military applications, many Deep Analytics ideas have been researched for years. For example, data fusion finds application in many military systems. Data fusion is often divided into a hierarchy of four processes. Level 1 and 2 fusion is generally concerned with processing raw data using numerical fusion methods. Level 3 and 4 fusion is thus concerned with the extraction of high-level knowledge from low level fusions, the incorporation of human judgment and the formulation of decisions and actions. Can these methods be Deep Models? Can these methods be adapted to the Big Data environment where data sources are heterogeneous and without standard data models or ontologies? How can ML and AI algorithms help Deep data fusion models?
There are many Deep Analytics and AI challenges for Military applications that have not been adequately addressed in industrial applications. Four of the main challenges include lack of adequate samples for classification tasks, short time scales for learning, less computational resources, and adversarial behavior. For example, suppose our task is to identify a rarely observed object for mission planning purposes. We may therefore have little training data, relatively little time to integrate recent observations into the training, only a network of high powered desktops to train, and any advisories may try to jam or corrupt our sensors. For these reasons, fast optimization methods, generative modeling, and transfer learning methods are of particular interest.
The objective of the workshop or working group is to foster collaborations and form communities for the theories and practices of Deep Models to military applications. We solicit unclassified research, papers, collaborative and innovative ideas in the following areas (not limited to) for military applications.
What are the potentials, theories, practices, tools and risks using the following Deep Models (i.e., models with large number of parameters that can be trained by Big Data)?
The workshop will consist of keynote talks, invited talks,
tutorials, oral presentations, poster/demo presentation, and panel discussions.
We will also invite a panel of experts from defense organizations, funding
agencies and contractors to discuss these topics as a part of the workshop
activities.
Regular papers should be 8 pages. Position papers should be 2 pages; submitted to https://easychair.org/conferences/?conf=fss17dmai4ma
Ying Zhao, Ph.D.
Research Professor
Information Sciences Department
Naval Postgraduate School
Monterey, CA 93943
yzhao@nps.edu
Arjuna Flenner (Navy-NAVAIR, China Lake)
Nate Derbinsky (Wentworth Institute of Technology)
David A. Bader (Georgia Institute of Technology)
Bonnie Johnson (Naval Postgraduate School)
Charles Zhou (Quantum Intelligence, Inc.)
Alexander Mezhirov (MIT Lincoln
Laboratory)
Alan (Ira) Clarke (Naval Postgraduate School)
Tom Starai (Navy Cyber Warfare Development Group,
NMIC)
Douglas MacKinnon (Naval Postgraduate School)
Shelley Gallup (Naval Postgraduate School)
Tony Kendall (Naval Postgraduate School)
Arkady Godin (Naval Postgraduate School)
Albert (Buddy) Barreto (Naval Postgraduate School)
Scott Humr (USMC)
William, Treadway (OPNAV)
Andrew Buckon (NAVSEA, PEO IWS)
Henry Salmans (USMC and Computer Sciences
Corporation)
Jerome Darbon (Brown University)
Rebecca Garnett (Navy-NAVAIR, China Lake)
Sean Shepherd (Navy-NAVAIR, China Lake)
Lawrence Peterson (Navy-NAVAIR, China Lake)
Levi Roberts (Navy-NAVAIR, China Lake)
Jesse Hodge (Navy-NAVAIR, China Lake)
Lavanya Iyer (Navy-NAVAIR, Pt. Mugu)
Daniel Omoto (Navy-NAVAIR, Pt. Mugu)
Ivan Tucker (Naval Nuclear
Laboratory)
John Reeder (SPAWAR, San Diego)
Joshua Harguess (SPAWAR,San
Diego)
http://faculty.nps.edu/yzhao/fss17-dmai4ma
AAAI 2017
Fall Symposium Series
(Deep Models and
Artificial Intelligence for Military Applications: Potentials, Theories,
Practices, Tools and Risks)
Thursday,
November 9, Session Chair: Dr. Ying Zhao, Naval
Postgraduate School, USA
9:00 am - 9:30 am: Invited Speaker: Ralucca Gera, Ph.D., Associate
Professor, Department of Applied Mathematics, Center for Cyber Warfare, Naval
Postgraduate School, USA
9:30
am – 10:00 am: Using D3 to Visualize Lexical Link Analysis (LLA) and ADS-B
Data, Quinn Halpin, Ying Zhao and Anthony Kendall, Cornell University, Naval
Postgraduate School, USA
10:00
am – 10:30 am: Rational Behavior Model (RBM) Architecture and Human-Robot
Ethical Constraints Using Mission Execution Ontology (MEO), Don Brutzman,
Curtis Blais and Robert McGhee, Naval Postgraduate School, USA
10:30
am – 11:00 am: break
11:00
am – 11:30 am: Analysis of Automatic Dependent Surveillance-Broadcast Data,
Ryan Salcido, Ying Zhao and Anthony Kendall, Naval Postgraduate School, USA
11:30
am - 12:00 pm: Panel discussion, lead -- Dr. Ying Zhao, Naval Postgraduate
School
12:30
pm - 2:00 pm: Lunch
2:00
pm – 2:30 pm A Systems Approach to Battle Management Aids, Bonnie Johnson,
Naval Postgraduate School, USA
2:30 pm – 3:00 pm A Framework Using Machine Vision and Deep
Reinforcement Learning for Self-learning Moving Objects in a Virtual
Environment, Richard Wu, Alan (Ira) Clarke, Ying Zhao, Anthony Kendall, Naval
Postgraduate School, USA
3:00
pm - 3:30 pm Sandia’s
Video Analytics capabilities, Wallace Bow, Military Systems & Technologies,
Sandia National Laboratories
3:30
pm - 4:00 pm Break
4:00
pm – 5:30 pm Panel discussion: lead -- Professor Anthony
Kendall, Naval Postgraduate School, USA
6:00
pm - 7:00 pm Reception
Friday, November 10, Session Chair: Professor Anthony Kendall, Naval Postgraduate School, USA
9:00
am - 9:30 am: Invited Speaker: Mr. William Treadway, Deputy Director, OPNAV
N2/N6 F33 Integrated Fires
9:30
am – 10:00 am:
10:00
am – 10: 30 am: Use Machine Learning To Design Better Security Vulnerability
Metrics For A Wireless IoT System, Philip Chan, Jessie Zhao, Samuel Chan,
UMUC, Stratford University, Duke University
10:30 am -
11:00 am Break
11:00
am - 11:30 pm Phylogenetic-Inspired Probabilistic Model Abstraction in
Detection of Malware Families, Krishnendu Ghosh,
Jeffery Mills and Joseph Dorr, Miami University
11:30
am – 12:30 pm Panel discussion: lead – Dr. Alex Roesler,
Deputy Director, Integrated Military Systems Military Systems &
Technologies, Sandia National Laboratories
12:30
pm - 2:00 pm Lunch
2:00
pm - 2:30 pm Towards a Repeated Bayesian Stackelberg
Game Model for Robustness Against Adversarial Learning. Prithviraj Dasgupta, Joseph Collins , University of
Nebraska, Omaha, Naval Research Lab, Washington DC
2:30pm - 3:00 pm Integration of Graphs and Representation Learning.
Arjuna Flenner, NAVAIR
3:00 pm - 3:30 pm: Simple Object Classification using Binary Data.
Deanna Needell, Rayan Saab and Tina Woolf, UCLA, UCSD and JPL, USA
3:30
pm – 4:30 pm: break
4:30
pm – 5:30 pm: Panel discussion: lead -- Dr. Arjuna Flenner, NAVAIR
6:00 pm - 7:00 pm: Plenary Session, speaker for our workshop: Mr.
William Treadway, Deputy Director, OPNAV N2/N6 F33 Integrated Fires
Saturday, November 11, Session Chair: Dr. Arjuna Flenner, NAVAIR
9:00 am - 9:30 am: Invited Speaker: Mr. Roshan Punnoose,
Technical Lead, Enlighten IT Consulting (EITC), MacAulay-Brown, Inc.
9:30 am – 10:30 am Autonomous Outcomes: Shaping the Future Data
Environment to Build Trust in Artificial Intelligence and Machine Learning
Applications. Maj. Scott Humr Marine Corps University
10:30
am – 11:00 am:
Break
11:00
am – 11:30 am:
Panel Discussion: lead – Maj. Scott Humr Marine Corps
University
11:30
am – 12:30 am:
Symposium
concludes at 12:30 pm
The AAAI 2017 Fall Symposium Link: https://aaai.org/Symposia/Fall/fss17symposia.php#fs03
Technical Report: http://www.aaai.org/Library/Symposia/Fall/fs17-03.php
The AAAI AI Magazine 2018 Special
Issue Volume
1 and Volume
2