Research Experience for Undergraduate Studies

Research Projects

Research Projects

Project 1: Machine Learning-based Energy-efficient Workload Management on Cluster Systems

Faculty Mentors: Dr. Xunfei Jiang and Dr. Mahdi Ebrahimi

This project will use machine learning techniques to develop a set of energy consumption models that simulate the energy cost of components in a cluster system and validate the models on real applications. REU students will develop a literature review to gain knowledge and experience of DS and PDC systems. They will conduct experiments to characterize energy consumption patterns of major components in the test bed cluster through data collection and analysis. Energy-efficient workload management and data placement strategies will be investigated by students, who will conduct experiments to analyze the energy efficiency of these strategies by considering task execution time and/or cost. In this project, students will gain immersive research experience on problem-solving through analyzing existing research on workload scheduling algorithms, proposing energy-efficient workload scheduling algorithms, designing and conducting experiments, and applying DS in data collection, data processing, and data analysis to evaluate the proposed solutions. They will learn machine learning skills and apply them on energy modeling and workload management on cluster systems, and work with group members through collaborative discussions with participants possessing diverse technical backgrounds and research preparation.

Project 2: Energy-efficient Geo-Information Visualization and User Interaction

Faculty Mentor: Dr. Li Liu and Dr. Mario Giraldo

Participant students will leverage energy-saving in: 1) designing geographic literacy and map reading approaches that are intuitive to users and their ability to understand weather information from a rich set of geographic data; and 2) evaluating the effect of abstraction and perceptualization of large amounts of complex, inherently non-spatial and unstructured forecasting data on a mobile user’s decision-making or behavioral change in emergency situations with the help of a cluster system. Students will visualize the information to be presented to stakeholders and build a user interface that provides easily accessible user interaction. In this project, REU students will gain experience of data processing, data visualization. They will collaborate with group members to integrate knowledge of DS and PDC on solving geospatial data visualization and user interaction problems.

Project 3: Energy Prediction for Automobile Air Conditioning Systems

Faculty Mentor: Dr. Taehyung Wang

Energy Prediction for Automobile Air Conditioning Systems will study the energy consumption of automobile air conditioning systems, which is a data-intensive application. The energy cost of automobile air conditioning systems is affected by many factors (such as external environment, driving environment, noises, etc.). This brings up a challenge for predicting the energy consumption of automobile air conditioning systems by analyzing correlated data. Students will work on over-the-road data to classify the amount of energy consumption and to find strong associations between features of the data. Students will gain knowledge and research experience in machine learning and deep learning through various pre-processing, modeling, and evaluation techniques to build high-performance and efficient models. Students will also tune the models to find an optimal model that minimizes a predefined loss function on given data using hyperparameter optimization.

Project 4: Coding-Based Data Storage for Optimizing Durability and Energy Efficiency in Phase Change Memory

Faculty Mentor: Dr. Marjan Asadinia

The next generation of memory technologies, like Phase Change Memories (PCMs), will address gaps in today’s storage hierarchy, delivering data where it’s needed for real-time processing. PCM promises to bridge the gap between DRAM and flash in the data center. This project proposes to explore efficient computer algorithms and architecture to design and implement phase change memory (PCM) to meet the storage needs of consumers. PCM has significant shortcomings compared to Dynamic RAM (DRAM). To overcome these challenges, we will improve the write energy, reliability, and durability of phase change memories, which are a key enabling technology for non-volatile electrical data storage at the nanometer scale. In this research, problems will be addressed from both a hardware and software perspective. From a hardware perspective, all proposed solutions aim to improve the write energy of a PCM main memory system while addressing the limited endurance, and hard errors problems. From a software perspective, it will involve operating system support for address translation from virtual memory address to real PCM address. Another facet of this project considers the design and development of Deep Neural Networks (DNNs) at a chip level. We exploit the characteristics of flash memory cells to create an efficient DNN within a chip. This can be achieved by either leveraging the structure of NAND flash memory cell strings (with separate word-line control) or adopting the structure of NOR flash memory cells (with shared word-line control). REU students will encompass conducting comprehensive literature reviews, contributing to the implementation and evaluation of proposed methodologies, and playing a role in manuscript preparation and presentation development. This opportunity offers valuable hands-on experience in research techniques, exposure to scientific tools, enhanced skills in scientific writing and presentation, a deeper comprehension of advanced memory technologies, and a broader understanding of deep neural networks (DNNs).