Time Series Analysis using a Kernel based Uncertainty Decomposition Framework
Conference on Uncertainty in Artificial Intelligence (UAI) 2020:
Composite Dynamic Texture Synthesis Using Hierarchical Linear Dynamical System
International Conference on Acoustics, Speech and Signal Processing (ICASSP) 2020:
Contributed Talks:
A Quantum Theory Inspired Framework for Uncertainty Quantification
IEEE International Conference on Data Science and Advanced Analytics (DSAA)
Speaker: Dr. Jose C. Principe
Abstract
This talk presents our current goal of developing operators inspired by quantum theory to quantify uncertainty in time series and train adaptive models for machine learning. The basic observation is that data projected to a Reproducing Kernel Hilbert Space (RKHS) are functions that obey the properties of a potential field. Therefore, one can directly apply the Schrodinger equation to the projected data and interpret its Hermite expansion in terms of modes over the space of samples that express multi scale uncertainty. This methodology can be used to quantify signal properties and can also lead to methodologies to train signal processing models. We will exemplify the theory with some preliminary results.
Microsoft Presentation
Speaker: Dr. Shujian Yu