Katherine JohnsonUsing groundbreaking artificial intelligence (AI) technology, a University of Kentucky researcher is developing a machine learning pipeline with the goal of improving our quality of life.
Hana Khamfroush, Ph.D., associate professor in the Department of Computer Science in the UK Stanley and Karen Pigman College of Engineering, recently received the prestigious National Science Foundation (NSF) Faculty Early Career Development (CAREER) Award. The NSF will support Khamfroush with $624,716 over five years for her research involving pre-processing data, while maintaining privacy, so that it can be trained for use in machine learning models for smart cities applications.
With eco-friendly practices as a priority, smart cities use data and technology to create more livable and sustainable urban environments.
“I think we are all used to the internet on computers and smartphones. But when we talk about the ‘internet of things,’ we are looking at every possible device becoming connected devices to the internet,” said Khamfroush. “For example, we can have a smart thermometer that can just sense when we are out of the home to reduce the lights. This can help with energy consumption.”
The NSF-funded work will serve as a foundation for various emerging AI-based applications including smart traffic light systems. Many of these applications will require a huge amount of data to be automatically processed and some will need to be processed in real time.
“There is a lot of noisy data and missing data points,” said Khamfroush. “A big part of this project is about federated learning and federated data preparation. This means we are preparing data and training machine learning models without losing privacy because we are not sharing the data to a cloud. All the training is done collaboratively and locally on the devices.”
Khamfroush’s research was previously focused on distributed and edge computing systems. As machine learning becomes more and more developed, she says her research becomes more applicable in the domain of machine learning and distributed machine learning.
