Feature selection is a critical step in the machine learning (ML) model development workflow, aimed at identifying the most relevant subset of features from a dataset to improve ML model performance. In this paper, we investigate the use of quantum annealing to enhance the efficiency and effectiveness of feature selection, as compared to classical algorithmic methods for feature selection, prior to constructing a ML model for Internet of Things network intrusion detection. We aim to determine the optimal selection of network traffic features that contribute most to the detection of network intrusions. Leveraging a quantum annealing algorithm, which exploits quantum mechanics principles to find optimal solutions, along with D-Wave’s hybrid quantum computing service, enables us to successfully tackle this combinatorial optimization problem. Our quantum machine learning approach leverages the strengths of both classical and quantum computing, offering a “one-shot” solution to feature selection without the need for iterative ML model training or incremental construction of the solution.
It is a major challenge to maintain Differential Privacy (DP) when fine-tuning a Large Language Model (LLM) while also preserving the increased functionality that make fine-tuned LLMs appealing. In this paper, we explore the utilization of an LLM that has been modified for the purpose of encoded message transmission using text output as a medium, a task which falls under the classification of Linguistic Steganography. By examining the impact that DP preserving fine-tuning has on an LLM intended for such a specific and technical functionality, we evaluate what performance cost is imparted. Our experimentation focuses on using a modified implementation of Differentially Private Stochastic Gradient Descent, while fine-tuning a LLM on curated data taken from the ConvoKit Reddit dataset. We were able to securely fine-tune the LLM while maintaining a relatively strict DP privacy budget, and still benefit from the domain specific increased performance that LLM fine-tuning provides.
Representing context, reasoning within contexts, and providing quantitative assessments of machine learning (ML) model certainty are all tasks of fundamental importance for secure, interpretable, and reliable model development. Recent enthusiasm regarding generative ML models has highlighted the importance of representing context, which is contingent on relevant and contextual features of data and model predictions are unreliable on out-of-context inputs. Herein, we develop the theory of graph representation learning (GRL) to extend to Bayesian Graph Neural Networks and to incorporate various forms of uncertainty quantification to improve model development and application in the presence of adversarial attacks. Within this framework, we approach the challenge of adversarial patch detection using a synthesized dataset consisting of images from the APRICOT and COCO datasets to study various binary classification models for patch detection. We present GRL models with two layers of edge convolution that are capable of detecting patches with up to 93.5% accuracy. Further, we find evidence supporting the use of the certainty and competence framework for model predictions as a tool for detecting patches, particularly when the former is included as a model feature in graph neural networks.
In tactical edge networks, the volatility of computational and communication resources complicates the consistent processing of data. Previous work has developed a system that allows execution of inference task applications throughout a network using a variety of adaptations to machine learning models that offer accuracy and latency trade-offs as conditions change in order adaptively perform deterministic resource allocation at the tactical edge. This paper expands on previous work that proposed a system that allows execution of inference task applications throughout a network and then developed a resource allocation algorithm that optimally places intelligence, surveillance, and reconnaissance related machine learning tasks throughout the network. We propose utilizing stochastic optimization to analyze the computational time and performance for inference tasks. Instead of adhering to deterministic averages, we use sample average approximation as a solution technique to optimize and analyze the inherent uncertainties of tactical edge environments, optimizing over the distribution of inference data rather than average-case scenarios. This paper verifies Jensen’s inequality gap within deterministic optimization and proposes an improved resource allocation algorithm that optimally places tasks throughout a network under uncertainty. We present initial results on a military relevant tactical edge network scenario.
Deep learning (DL) has revolutionized machine learning tasks in various domains, but conventional DL methods often demand substantial amounts of labeled data. Semi-supervised learning (SSL) provides an effective solution by incorporating unlabeled data, offering significant advantages in terms of cost and data accessibility. While DL has shown promise with its integration as a component of modern network intrusion detection systems (NIDS), the majority of research in this field focuses on fully supervised learning. However, more recent SSL algorithms leveraging data augmentations do not perform optimally “out of the box” due to the absence of suitable augmentation schemes for packet-level network traffic data. Through the introduction of a novel data augmentation scheme tailored to packet-level network traffic datasets, this paper presents a comprehensive analysis of multiple SSL algorithms for multi-class network traffic detection in a few-shot learning scenario. We find that even relatively simple approaches like vanilla pseudo-labeling can achieve an F1-Score that is within 5% of fully supervised learning methods while utilizing less than 2% of the labeled data.
Artificial intelligence (AI) is quickly gaining relevance as a transformative technology. Its ability to rapidly fuse and synthesize data, accelerate processes, automate tasks, and augment decision-making has the potential to revolutionize multi-domain warfighting through data-centric operations and algorithmic warfare. As the military relies more on AI-enabled Decision Aids to increase the efficiency and effectiveness of decision-making, it highlights the need to effectively assess them before deployment. Modeling and simulation (M&S) environments are essential for assessing these rapidly evolving AI-enabled systems. Accepted analytical frameworks are needed to guide ways to represent and model AI sufficiently within M&S environments for accurate assessment. In this paper, we identify common characteristics within the main categories of AI and investigate how those characteristics can be best represented across the main categories of M&S. We provide two use cases to highlight an assessment of AI-enabled Decision Aids for cybersecurity and aeromedical evacuation problems. Our example use cases demonstrate how to leverage a framework for analytic assessment of AI within M&S environments.
In recognizing the importance of network traffic monitoring for cybersecurity, it is essential to acknowledge that most traditional machine learning models integrated in network intrusion detection systems encounter difficulty in training because acquiring labeled data involves an expensive and time-consuming process. This triggers an in-depth analysis into zero-shot learning techniques specifically designed for raw network traffic detection. Our innovative approach uses clustering combined with the instance-based method for zero-shot learning, enabling classification of network traffic without explicit training on labeled attack data and produces pseudo-labels for unlabeled data. This approach enables the development of accurate models with minimal limited labeled data for making network security more adaptable. Extensive computational experimentation is performed to evaluate our zero-shot learning approach using a real-world network traffic detection dataset. Finally, we offer insights into state-of-art developments and guiding efforts to enhance network security against ever-evolving cyber threats.
KEYWORDS: Network security, Machine learning, Defense and security, Computer networks, Monte Carlo methods, Inspection, Data modeling, Windows, Transformers
In this work, we demonstrate the potential of dynamic reinforcement learning (RL) methods to revolutionize cybersecurity. The RL framework we develop is shown to be capable of shutting down an aggressive botnet, which initially uses spear phishing to establish itself in a Department of Defense (DoD) network. To ensure a suitable real-time response, we employ CP, a transformer model trained for network anomaly detection, to factorize the state space accessible to our RL agent. As the fidelity of our cyber scenario is of the utmost importance for meaningful RL training, we leverage the CyberVAN emulation environment to model an appropriate DoD enterprise network to attack and defend. Our work represents an important step towards harnessing the power of RL to automate general and fully-realistic Defensive Cyber Operations (DCOs).
KEYWORDS: Artificial intelligence, Data modeling, Decision making, Risk assessment, Data fusion, Systems modeling, Network security, Sensors, Safety, Information fusion
Many techniques have been developed for sensor and information fusion, machine and deep learning, as well as data and machine analytics. Currently, many groups are exploring methods for human-machine teaming using saliency and heat maps, explainable and interpretable artificial intelligence, as well as user-defined interfaces. However, there is still a need for standard metrics for test and evaluation of systems utilizing artificial intelligence (AI), such as deep learning (DL), to support the AI principles. In this paper, we explore the elements associated with the opportunities and challenges emerging from designing, testing, and evaluating such future systems. The paper highlights the MAST (multi-attribute scorecard table), and more specifically the MAST criteria ―analysis of alternatives‖ by measuring the risk associated with an evidential DL-based decision. The concept of risk includes the probability of a decision as well as the severity of the choice, from which there is also a need for an uncertainty bound on the decision choice which the paper postulates a risk bound. Notional analysis for a cyber networked system is presented to guide to interactive process for test and evaluation to support the certification of AI systems as to the decision risk for a human-machine system that includes analysis from both the DL method and a user.
Machine learning (ML) requires both quantity and variety of examples in order to learn generalizable patterns. In cybersecurity, labeling network packets is a tedious and difficult task. This leads to insufficient labeled datasets of network packets for training ML-based Network Intrusion Detection Systems (NIDS) to detect malicious intrusions. Furthermore, benign network traffic and malicious cyber attacks are always evolving and changing, meaning that the existing datasets quickly become obsolete. We investigate generative ML modeling for network packet synthetic data generation/augmentation to improve NIDS detection of novel, but similar, cyber attacks by generating well-labeled synthetic network traffic. We develop a Cyber Creative Generative Adversarial Network (CCGAN), inspired by previous generative modeling to create new art styles from existing art images, trained on existing NIDS datasets in order to generate new synthetic network packets. The goal is to create network packet payloads that appear malicious but from different distributions than the original cyber attack classes. We use these new synthetic malicious payloads to augment the training of a ML-based NIDS to evaluate whether it is better at correctly identifying whole classes of real malicious packet payloads that were held-out during classifier training. Results show that data augmentation from CCGAN can increase a NIDS baseline accuracy on a novel malicious class from 79% to 97% with a minimal degradation in accuracy on benign classes (98.9% to 98.7%).
In this work, we aim to develop novel cybersecurity playbooks by exploiting dynamic reinforcement learning (RL) methods to close holes in the attack surface left open by the traditional signature-based approach to Defensive Cyber Operations (DCO). A useful first proof-of-concept is provided by the problem of training a scanning defense agent using RL; as a first line of defense, it is important to protect sensitive networks from network mapping tools. To address this challenge, we developed a hierarchical, Monte Carlo-based RL framework for the training of an autonomous agent which detects and reports the presence of Nmap scans in near real-time, efficiently and with near-perfect accuracy. Our algorithm is powered by a reduction of the state space given by a transformer, CLAPBAC, an anomaly detection tool which applies natural language processing to cybersecurity in a manner consistent with state-of-the-art. In a realistic scenario emulated in CyberVAN, our approach generates optimized playbooks for effective defense against malicious insiders inappropriately probing sensitive networks.
Traditional machine learning (ML) models used for enterprise network intrusion detection systems (NIDS) typically rely on vast amounts of centralized data with expertly engineered features. Previous work, however, has shown the feasibility of using deep learning (DL) to detect malicious activity on raw network traffic payloads rather than engineered features at the edge, which is necessary for tactical military environments. In the future Internet of Battlefield Things (IoBT), the military will find itself in multiple environments with disconnected networks spread across the battlefield. These resource-constrained, data-limited networks require distributed and collaborative ML/DL models for inference that are continually trained both locally, using data from each separate tactical edge network, and then globally in order to learn and detect malicious activity represented across the multiple networks in a collaborative fashion. Federated Learning (FL), a collaborative paradigm which updates and distributes a global model through local model weight aggregation, provides a solution to train ML/DL models in NIDS utilizing learning from multiple edge devices from the disparate networks without the sharing of raw data. We develop and experiment with a data-efficient, FL framework for IoBT settings for intrusion detection using only raw network traffic in restricted, resource-limited environments. Our results indicate that regardless of the DL model architecture used on edge devices, the Federated Averaging FL algorithm achieved over 93% accuracy in model performance in detecting malicious payloads after only five episodes of FL training.
Detecting malicious activity using a network intrusion detection system (NIDS) is an ongoing battle for the cyber defender. Increasingly, cyber-attacks are sophisticated and occur rapidly, necessitating the use of machine/deep learning (ML/DL) techniques for network intrusion detection. Traditional ML/DL techniques for NIDS classifiers, however, are often unable to sufficiently find context-driven similarities between the various network flows and/or packet captures. In this work, we leverage graph representation learning (GRL) techniques to successfully detect adversarial intrusions by exploiting the graph structure of NIDS data to derive context awareness, as graphs are a universal language for describing entities and their relationships. We explore several methods for NIDS data graph representation at both the network flow and packet level utilizing the CIC-IDS2017 dataset. We leverage graph neural networks and graph embedding algorithms to create a context-aware network intrusion detection system. Results indicate that adding context derived from GRL improves performance for detecting attacks. Our highest-scoring classifier incorporated both GNN embeddings and flow-level features and achieved an accuracy of 99.9%. Adding GRL methods to augment the flow/packet features improved accuracy by as much as 52.41%.
As methods and access to gene synthesis and genetic engineering have become more advanced, the fear that malicious viruses and bacteria will be designed with the express intention of causing harm to humans has received increased attention. In the event that such biological weapons are deployed, the security community needs tools to rapidly recognize the threat and identify responsible parties. Therefore, a key question is whether or not a biological threat is manmade. Currently, experts are capable of qualitatively assessing whether specific genetic sequences are natural or man-made, but few objective criteria exist for characterizing the degree to which a sequence has been engineered. Additionally, progress has recently been made on the task of attributing an engineered gene sequence to a lab-of-origin using machine learning. However, the task of analyzing naturally occurring genetic sequences so as to automatically detect outliers that may have been genetically engineered has received comparatively little attention. This work proposes a method for generating a dataset of natural and engineered sequences that can be used as an input for training machine learning classifiers to perform automatic detection of human engineering in gene sequence data.
Increasingly cyber-attacks are sophisticated and occur rapidly, necessitating the use of machine learning techniques for detection at machine speed. However, the use of machine learning techniques in cyber security requires the extraction of features from the raw network traffic. Thus, subject matter expertise is essential to analyze the network traffic and extract optimum features to detect a cyber-attack.
Consequently, we propose a novel machine learning algorithm for malicious network traffic detection using only the bytes of the raw network traffic. The feature vector in our machine learning method is a structure containing the headers and a variable number of payload bytes. We propose a 1D-Convolutional Neural Network (1D-CNN) and Feed Forward Network for detection of malicious packets using raw network bytes.
In order to scale for speed, technology often builds upon the earliest proven systems and architectures. As the context changes, from a civilian application domain to a military application domain, the priority of functional requirements can and often do change. The hardware, software, and language development environment set the foundation for the constraints and potential of a system. This along with the fact the information technology revolution, since early 2000, has primarily been driven by the commercial sector, requires engineers to consider whether nontraditional, less well-known architectures may have a role in the Multi-Domain Operations (MDO) application space. This paper will highlight features inherent to traditional architectures, the challenges associated with these architectural features, and how the Erlang VM represents an opportunity to develop an architectural foundation suitable to the MDO application domain. Finally, this paper will highlight a future technology concept integrating demonstrated neural interface technology with an Erlang VM supported architecture. This foundation will help enable human-machine teaming by empowering a human agent to interact with sensors and AI-enabled autonomous systems with a dynamic user interface allowing the human agent to accomplish MDO applications. The great potential for the concept depends on a fault-tolerant, distributed system permitted by the Erlang VM to exibly integrate the capabilities required to address the diverse challenges of a complex operating environment.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.