Machine Learning Optimization & Signal Processing Laboratory

The RIT Machine Learning Optimization & Signal Processing Laboratory (MILOS Lab) was founded in 2016 by Dr. Panos Markopoulos as a team of researchers exploring the areas of machine learning, data science, and adaptive signal processing, with an aim to advance efficient, explainable, and trustworthy artificial intelligence.

Our mission has three main pillars:

• Conduct fundamental research in the areas of machine learning, data analysis, and adaptive signal processing.

• Deliver solutions to technological challenges that address societal needs and are of interest to the nation.

• Provide top-quality inclusive education and train the next generation of innovators and technology leaders.

Our expertise is in the areas of machine learning, data science, and signal processing, with a focus on both theoretical (computational and statistical) foundations and on practical algorithmic solutions, evaluated in a wide range of real-world applications.

Current research topics include:

• Dynamic data subspace learning based on Lp-norm projections.

• Robust learning with limited, faulty, and adversarially corrupted data.

• Tensor methods for multi-way data analysis and processing.

• Tensor factorization methods for modeling efficient and explainable neural networks.

• Learning from multimodal data and deep learning fusion (recent topic).

• Continual learning with increasing parameter tensor ranks (recent topic).

Among multiple other areas, our fundamental research has found applications in remote sensing, computer vision, communication systems, and healthcare technology.

• National Science Foundation (NSF).

• Air Force Research Lab (AFLR), Information Directorate.

• Air Force Office of Scientific Research (AFOSR).

• National Geo-Spatial Intelligence Agency (NGA).

• NYSTAR / UR CoE in Data Science.

• L3Harris.

• RIT Office of the Vice President for Research (OVPR) and RIT Kate Gleason College of Engineering (KGCOE).