- 2021-2022 : Computer Science New Faculty Candidate Presen...
Posted 09:24 Feb 24, 2022 |
Computer Science New Faculty Candidate Presentations
The Department of Computer Science's new faculty candidates will present their research work. All are welcome. Come and meet them.
MD Ali Reza Al Amin
PhD Candidate, Old Dominion University
Monday, 2/28/2022, 10:00AM-11:00AM
Privacy Preserving Analytics to Enable Intrusion Detection Data Sharing
Supervised learning is effectively adopted in Network Intrusion Detection Systems (IDS) to detect malicious activities in Information Technology (IT) and Operation Technology (OT). Sharing high-quality network data among cyber-security practitioners increases the chance to detect new threat campaigns by analyzing updated traffic features. As data sharing brings privacy concerns, Differential-Privacy (DP) has emerged as a promising approach to perform privacy-preserving analytics. This research develops a framework to generate high-quality synthetic network features using differentially private Generative Adversarial Network (DP-GAN). A well-known intrusion detection dataset, NSL-KDD, is used to conduct the experiments. To date, NSL-KDD has still been considered an intrusion detection benchmark because of its diverse attacks groups. The experiment records the classification performance of several machine learning (ML) models on a privacy-preserved synthetic dataset derived from the NSL-KDD intrusion dataset.
PhD Candidate, Clemson University
Tuesday, 3/1/2022, 1PM-2PM
Internet for Internet of Things
An Internet of Things (IoT) system is an interconnected system of nodes that can collaborate to carry out common objectives through applications running on them. IoT systems are typically defined by wireless connectivity and resource constraints in computation, memory, communication, and battery power. Popular IoT systems include a Connected Vehicular Network and a swarm of Unmanned Aerial Vehicle (UAV swarm). These systems are currently being applied to many practical use-cases and especially fulfill essential roles in society’s critical infrastructure. This significant utility and economic viability of deploying these systems is a crucial indicator of their anticipated growth in the future. The applications running on these systems are also increasingly becoming more complex. Emerging applications for these systems require sensing, perception, and analysis of substantial amount data with strict performance requirements. Therefore, these systems require system connectivity, data dissemination, and data analysis methods that can operate well within their limitations. However, traditional Internet protocols and network connectivity methods that are designed for well-engineered cyber systems do not fully meet these requirements. The imminent growth of IoT systems presents an opportunity to develop broadly applicable solutions that can meet the IoT system and application requirements and better integrate these systems with the Internet.
In this talk, I will present my recent research contributions towards designing such a solution called the Software-Defined NAmed-data enables Publish-Subscribe (SNAP) framework. Through concepts leveraged from Software-Defined Networking (SDN) and Information-Centric Networking (ICN) paradigms, the SNAP framework provides robust and lightweight methods to support system connectivity and data dissemination driven by the IoT system’s characteristics and performance requirements. The SNAP framework can be inclusively applied to any IoT system and extended to support inter-operability among them. It also provides methods for the integration of IoT systems with the Internet.
Assistant Professor, Computer Science and Information System, Bradley University
Wednesday, 3/2/2022, 10:00AM-11:00AM
Improving Reliability and Durability of Phase Change Main Memories
With current memory scalability challenges, Phase-Change Memory (PCM) is viewed as an attractive replacement to DRAM. The preliminary concern for PCM applicability is its limited write endurance that results in fast wear-out of memory cells. Worse, process variation in the deep-nanometer regime increases the variation in cell lifetime, resulting in an early and sudden reduction in main memory capacity due to the wear-out of a few cells. Recent studies have proposed redirection or correction schemes to alleviate this problem, but all suffer poor throughput or latency. In this talk, I present that one of the inefficiency sources in current schemes, even when wear-leveling algorithms are used, is the nonuniform write endurance limit incurred by process variation, that is, when some memory pages have reached their endurance limit, other pages may be far from their limit. In this line, I propose a technique that aims to displace a faulty page to a healthy page. This technique, when applied at page level, can improve PCM Reliability and Durability.
PhD Candidate, University at Buffalo
Thursday, 3/3/2022, 1PM-2PM
Impact of Batch Refactoring Code Smells on Application Resource Consumption
Presentation abstract: Automated batch refactoring has become a de-facto mechanism to restructure software that may have significant design flaws negatively impacting the code quality and maintainability. Although automated batch refactoring techniques are known to significantly improve software quality and maintainability, their impact on resource utilization is not well studied. We investigate 16 code smell types and their joint effect on resource usage for 31 open source applications. We provide a detailed empirical analysis of the change in application CPU and memory usage after refactoring specific code smells in isolation and in batches. This analysis is then used to train regression algorithms to predict the impact of batch refactoring on CPU and memory utilization before making any refactoring decisions. Experimental results show that ANN-based regression model provides highly accurate predictions for the impact of batch refactoring on resource consumption. It allows the software developers to intelligently decide which code smells they should refactor jointly to achieve high code quality and maintainability without increasing the application resource utilization.
Research Scientist, Computer & Information Science, University of Florida
Monday, 3/7/2022, 10:00AM-11:00AM
AI Supported Learning for Computer Science
Adaptive and data-driven systems are changing the face of education and transforming traditional learning environments. From intelligent learning analytics to virtual pedagogical agents, AI applications present innovative and engaging opportunities to create personalized learning environments and support students as they pursue learning goals. With the growing focus on adaptive support for learners, developing computational models of the processes and phenomena in learners’ behaviors is becoming an increasingly important goal. In this talk, I will first present my research on the design, development and evaluation of an AI supported learning system that utilizes natural language processing (NLP), machine learning, and visualizations to empower learners to reflect on their collaborative dialogue and improve their collaborative behavior during pair programming activities. Results from our iterative user studies showed that after viewing their dialogue visualizations, learners engaged in more balanced dialogues, and even less-engaged students talked more and asked more questions. Next, I will expand on our recent research on computational approaches to the analysis of learners’ behaviors, including multimodal learning analytics, which provides new insights into students’ learning by analyzing multiple streams of data (e.g. speech, faces, and gestures) during a learning activity. Finally, I will discuss the growing momentum around AI learning, which points to promising future directions for bringing authentic, situated AI learning into classrooms through conversational AI (e.g., chatbots), NLP, and visualizations.
Associate Professor, Compute Science, Kettering University
Tuesday, 3/8/2022, 1PM-2PM
Vehicle Edge Computing
Connected and Autonomous Vehicles (CAV) are swiftly transforming the automotive industry, with emerging services driving major new requirements for data communications, which are expected to make the automotive sector the industry segment with the fastest growing demand for mobile machine-to-machine connectivity. The challenge is in designing and deploying the communication networks and computing ecosystem required to efficiently deliver and process the new high-volume data requirements. Considering the global nature of this challenge, we should consider how communication networks and computing resources could be orchestrated to enable secure, cost-effective data delivery and processing on a global scale. To address this challenge, Dr. Yunsheng Wang will share his recent work on “Vehicle Edge Computing” as a solution to ensure service flexibility, efficiency, and continuous evolution of the automotive industry.