— Non-negative Matrix factorization (NMF) has increasingly been used as a tool in signal processing in the last couple of years. NMF, like independent component analysis (ICA) is...
Variability of process parameters makes prediction of digital circuit timing characteristics an important and challenging problem in modern chip design. Recently, statistical stat...
Hongliang Chang, Vladimir Zolotov, Sambasivan Nara...
We suggest a recurrent neural network (RNN) model with a recurrent part corresponding to iterative function systems (IFS) introduced by Barnsley 1] as a fractal image compression ...
Most highly accurate predictive modeling techniques produce opaque models. When comprehensible models are required, rule extraction is sometimes used to generate a transparent mod...
Command-oriented functional programs are currently structured in an ad hoc way that makes the development of multiple userinterfaces difficult and error prone, and makes it diffic...