Ying-zong Huang

I received my Ph.D. degree in Electrical Engineering from MIT, where I wrote the following --


Dissertation:

I show why classic compression systems have a hard time making use of new knowledge about data, and how an entirely different design with a "model-code separation architecture" can do better. This won the Jin-Au Kong Thesis Award, Honorable Mention.

(There is also an associated software toolbox. For commercial concerns, inquire about technology licensing.)


Research:

I'm interested in information representation and statistical inference problems on complex signals, including neural-network representations. I am particularly fond of perceptual / multimedia signals, and I've designed algorithms for their optimal procesing, storage, and communication. More recently I was involved in building distributed algorithms and large-scale systems for applied machine learning.


Publications:

Machine Learning

Data Compression

Streaming Media

Classification


Unpublished results:


Notes:


Code:
MIT GitHub (requires auth)
GitHub


From a previous life:
Embedded systems
Spectral imaging
Linguistics
Genomics


More:
Signals, Information, and Algorithms Lab