I am an Applied Scientist at Amazon. I work on the development of Reliable Methods for Smart Home Automation. My research is broadly in the areas of Reliable Tensor Data Analysis, Robust Subspace Learning & Tracking, Deep Learning Model Compression, and Natural Language Processing. Recently, I obtained my PhD in Electrical and Computer Engineering from the Rochester Institute of Technology, Rochester, New York, USA, under the guidance of Dr. Panos P. Markopoulos.
Applied Scientist II, Amazon, August 2022 - Present.
Applied Scientist Intern, Amazon, May 2021 - August 2021. Performed research and developed methods based on Neural Architecture Search (NAS) and Tensor Methods for the development of lightweight CNNs that can be deployed on resource-constrained edge devices.
PhD Research Areas:
Formulation and development of theory and algorithms for Machine Learning and Signal Processing. Tensor processing methods for convolutional neural networks (CNNs ) compression and understanding. Specific areas include:
Outlier-resistant stochastic principal component analysis and adaptive subspace tracking.
Asymptotic understanding of L1-norm principal component analysis (L1-PCA).
Tensor processing for CNN compression, understanding, and better generalization.
Multi-modal data fusion for object detection in aerial imagery using convolutional neural networks.
Tensor based structured data fusion and learning in deep neural networks.