1-bit compressed sensing, sparse binary regression, and random hyperplane tessellations
Neyman Seminar
Jan 30, 2013, 04:00 PM - 05:00 PM | 1011 Evans Hall | Happening As Scheduled
Yaniv Plan, University of Michigan
Just as compressed sensing is closely related to linear model
selection, 1-bit compressed sensing is similar to sparse binary regression.
In the latter two problems we take the measurements from the former two
problems and then remove all information except for their signs. In 1-bit
compressed sensing this corresponds with an extreme level of quantization,
whereas in many statistical...