Creates a set of pattern (each later representing a mode) which than are randomly perturbed to create the data set. The dataset was introduced in Desjardins et al. (2010) (Parallel Tempering for training restricted Boltzmann machines, AISTATS 2010) More...
#include <shark/Unsupervised/RBM/Problems/DistantModes.h>
Public Member Functions | |
DistantModes (double p=0, unsigned dim=16, unsigned modes=4, unsigned copies=2500, size_t batchSize=0) | |
UnlabeledData< RealVector > | data () const |
returns the generated dataset More... | |
std::size_t | inputDimension () const |
returns the dimensionality of the data More... | |
Creates a set of pattern (each later representing a mode) which than are randomly perturbed to create the data set. The dataset was introduced in Desjardins et al. (2010) (Parallel Tempering for training restricted Boltzmann machines, AISTATS 2010)
The higher the perturbation is the harder it is to classify, but the closer are the modes and thus the easier the data distribution is to learn.
Definition at line 45 of file DistantModes.h.
|
inline |
generates the DistantModes distribution.
p | the probability of changing a input neuron |
dim | the dimensionality of the data. |
modes | the number of modes, should be a multiple of 2 |
copies | the number of disturbed copies for each mode |
batchSize | the size of the batches in which the generated data set is organized |
Definition at line 93 of file DistantModes.h.
|
inline |
returns the generated dataset
Definition at line 99 of file DistantModes.h.
|
inline |
returns the dimensionality of the data
Definition at line 104 of file DistantModes.h.