Implementation of sparse autoencoder in Matlab, based on the specifications of Stanford University's Machine Learning course. Training done on 8X8 image patches, using the L-BFGS optimization algorithm.
NOTE: Codebase is private. To request access, drop me a line at gautam @ bu . edu. Thanks!