Research
Extracting information from noise in physical measurements
I'm conducting independent research advised by Dr. Peter Beyersdorf at San José State University exploring uncertainty quantification methods based on noise statistics. The idea is that in any physical measurement, noise is unavoidable, and its statistical properties can be analyzed to assign calibrated confidence to inferences made from that measurement. Thermal noise is the cleanest test case because its behavior is well-characterized by physics, so any deviation from expected statistics is meaningful.
The work is structured in three phases: establishing statistical reference standards, detecting hidden structure through deviations from expected distributions, and developing uncertainty quantification methods for real-world inference like disease detection.
Learn more →A noisy MRI. The noise floor can't be eliminated, but its statistics may carry information about what's in the image.

