List view

• Future Features

  No due date

  •4/8 issues closed
  50% complete4 open4 closed

• 0.1beta

  Create the basic facilities for neural network prototyping, such as building blocks and topological operators. Implement the apply() and backpropagate() for them. Create the autoencoder mechanism for basic signal processing. Implement a basic version of full-batch training. Create documentation. -Jianbo Ye

  No due date

  •4/8 issues closed
  50% complete4 open4 closed

• A serial basic library

  I have prototyped such a template based neural network framework. Based on this, I have implemented a sparse auto-encoder for small size image patches (see data/UFLDL directory for bank of 8x8 patches). The serial performance is moderate, comparable to a vectorized Matlab implementation. Based on a JVM `-Xprof` profile, the major cost is calling libBLAS routines. In current version, the implementation is not fully exposed to breeze. I create one more interface layer for linear algebra (vector and matrix), such that I can possibly change to another numerical library.

  No due date

  •2/2 issues closed
  100% complete0 open2 closed