Research Topics

Below is a list of research topics supported by the AFRL. Use the filters and keyword search below to find research topics of interest. You can apply for up to 3 topics on your application.




Scholars are encouraged to contact any mentors whose projects they find of interest. To contact the mentor, use the link included at the conclusion of each project description.

A Network-based Approach on Modeling the Adversary’s Decision Calculus
Mentor: Jessica P Dorismond, Information
Location: Rome
Academic Level: Masters, Ph.D., Upper-level Undergraduate

A model of an adversaries’ decision calculus can be constructed to inform the development of adaptive strategies, including strategies that include effects-based operations. This topic will use a network theory approach to explore the adversary’s decision calculus. This network-based approach allows the tools of network analysis and graph theory to be brought together to analyze the overall topology of the decision network and to identify the key decision nodes and domains of outcomes associated with an adversary’s decision making.

Contact mentor

Characterizing Resilience of Distributed Systems
Mentor: Jeffrey Hudack, Information
Location: Rome
Academic Level: Masters, Ph.D.

The trend towards distributed systems provides benefits for cost, simplicity, and increased tolerance for disruption. However, as interactions are required between component systems these benefits come at the cost of increasing complexity of planning for deployment and operation. This project will focus on strategies for characterizing resilience of distributed systems, with potential projects focused on one or more of the following topics: 1) designing resilient systems, balancing trade-offs such as cost and robustness, 2) managing spatial and temporal dependencies among distributed systems, and their effects on resilience, 3) over-provision of resources to maximize resilience over time, providing enduring capability in the midst of disruptions.

Contact mentor

Data Efficient Machine Learning
Mentor: Walter D Bennette, Information
Location: Rome
Academic Level: Masters, Ph.D., Upper-level Undergraduate

A trained classification model can be used to automatically sort items of interest, such as images or signals, into discrete categories or classes. Modern machine learning algorithms typically require a vast amount of labeled data with which to learn such a classification scheme. Unfortunately, this quantity of data is sometimes difficult to acquire. We are interested in techniques that help users efficiently label datasets to reduce the amount of labeled data required to achieve a useful classification model. Your work will focus on implementing state of the art techniques for data efficient machine learning, running simulation experiments, and helping develop the next generation of efficient labeling procedures. Other projects will be introduced over the summer with opportunities for additional involvement.

Contact mentor

Game Theoretical Insights Into the Competition Phase of Multi-Domain Warfare
Mentor: David Myers, Information
Location: Rome
Academic Level: Masters, Ph.D., Upper-level Undergraduate

Game theory enables the development of important insights for interacting agents or decision-makers. Extensive research exists to understand various game-types, to include cooperative/non-cooperative, simultaneous/sequential, and perfect information/imperfect information. This effort will explore the modeling of a competition as a multi-stage, multi-player, multi-domain game where there are competing players looking to best position themselves against adversaries across multiple domains (through investments, positioning of capabilities, etc). This effort should focus on the game formulation, and any possible analysis conducted on the formulation of the game with the generation of simulated data. Insights of interest include the impact of non-competitive players, the impact of successful pre-positioning capabilities, among many others.

Possible considerations for other research extensions include the exploration of network flow under interdiction attacks (positioning of capabilities in a network) through the exploitation of the existing network or generation of new components of the network (vehicles, routes, nodes, edges, etc.).

Contact mentor

Generalizing from Few Examples with Meta Machine Learning
Mentor: Matthew M Klawonn, Information
Location: Rome
Academic Level: Masters, Ph.D.

State of the art machine learning models have achieved good generalization performance in a range of application domains, such as computer vision, natural language processing, etc. This is surprising because these models often contain an order of magnitude more trainable parameters than there are available data points. According to traditional ML wisdom, such models should overfit the training data. This topic proposes to investigate emerging theories explaining why overfitting does NOT occur in such overparameterized models. In doing so, we hope to uncover insights that will allow for learning in domains with highly limited training data. Our intuition is that meta-learning, i.e learning the learning process, will play a role in a data efficient learning regime.

Contact mentor

Interactive Machine Learning
Mentor: Matthew M Klawonn, Information
Location: Rome
Academic Level: Masters, Ph.D., Upper-level Undergraduate

Many recent efforts in machine learning have focused on learning from massive amounts of data resulting in large advancements in machine learning capabilities and applications. However, in certain domains obtaining explicit labels for training data can be challenging. In such cases, it is often desirable to collect feedback from subject matter experts via queries that elicit implicit information. For example, when ranking items as is done in search algorithms, it is often easier to ask a user to compare two items than to assign each a number in isolation, and rank them later. Posing such queries and learning from such feedback falls into a subfield of machine learning called interactive learning. Scholars will be asked, with help, to read and understand materials describing existing interactive learning techniques, and apply them to novel domains that may include, multi-objective optimization problems, , generative models, and more.

Contact mentor

Planning for Wargaming
Mentor: Brayden Daniel Hollis, Information
Location: Rome
Academic Level: Masters, Ph.D., Upper-level Undergraduate

This topic seeks to develop automated planners that can intelligently achieve missions in adversarial environments. Wargames have long been an effective method for developing and validating tactics, but traditionally rely on human players to provide sufficient challenge and variety. Students’ work may include encoding wargames into machine readable formats, implementing existing automated planning techniques to develop intelligent A.I. opponents, and developing new approaches for planning against an adversary. Possible extensions also include interactive planning techniques that allow automated planners to work with humans to develop robust plans.

Contact mentor

Reinforcement Learning and Planning for Trustworthy Autonomy
Mentor: Alvaro Velasquez, Information
Location: Rome
Academic Level: Masters, Ph.D., Upper-level Undergraduate

Reinforcement learning and sequential decision-making have been revolutionized in recent years thanks to advancements in deep neural networks. One of the most recent breakthroughs was accomplished by the AlphaGo system and its victory over the world Go champion. However, even in this impressive system, the learned agent performed sub-optimal actions that puzzled both the Go and the reinforcement learning communities. Such failures in decision-making motivate the need for methods that can provide (statistical) guarantees on the actions performed by an agent. Under this topic, we will explore novel decision-making techniques from the model-free perspective of reinforcement learning as well as from model-based views of control theory and planning. As a second technical aim, we will integrate statistical guarantees on the behavior of our learned agents. This is an open-ended research endeavor that seeks to provide a minimally intrusive direction on the students’ own intellectual proclivities. Since reinforcement learning and decision-making are highly interdisciplinary topics, the idea is for candidates to leverage their own knowledge and interests on this set of highly relevant problems.

Contact mentor

Semantic Machine Learning
Mentor: Patrick John Fisher, Information
Location: Rome
Academic Level: Masters, Upper-level Undergraduate

As complexity across domains increases and missions become more difficult, the need to become an interoperable military is necessary to maintain advantage. In the past, our knowledge and work force were able to out maneuver and dominate in the multi-domain spectrum of land, air, and sea. Today, we are slowly losing the ability to keep up with the overwhelming amount information coming from joint all domain operations. In order to help understand, assimilate, and bridge the gap between gathering the data and then wrangling said data into usable information, this project will focus on the following technologies: 1) leveraging state-of-the-art machine learning algorithms to assimilate and understand information from disparate datasets, 2) applying knowledge graphs to organize information across joint all domains, so that unanticipated questions can be answered for mission needs at a moment’s notice, 3) using visualization to uncover and understand the important parts of the information pipeline, so that answers can be conveyed.

Contact mentor