Home | Trees | Indices | Help |
|
---|
|
Perform a Hessian Locally Linear Embedding analysis on the data. Internal variables of interest: self.training_projection -- the HLLE projection of the training data (defined when training finishes) self.desired_variance -- variance limit used to compute intrinsic dimensionality Implementation based on algorithm outlined in Donoho, D. L., and Grimes, C., Hessian Eigenmaps: new locally linear embedding techniques for high-dimensional data, Proceedings of the National Academy of Sciences 100(10): 5591-5596, 2003. Original code contributed by: Jake Vanderplas, University of Washington vanderplas@astro.washington.edu
|
|||
Inherited from Node | |||
---|---|---|---|
__metaclass__ This Metaclass is meant to overwrite doc strings of methods like execute, stop_training, inverse with the ones defined in the corresponding private methods _execute, _stop_training, _inverse, etc... |
|
|||
|
|||
|
|||
Inherited from |
|||
Inherited from LLENode | |||
---|---|---|---|
|
|||
|
|||
|
|||
|
|||
|
|||
Inherited from Cumulator | |||
|
|||
|
|||
|
|||
Inherited from Node | |||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|
|||
Inherited from |
|||
Inherited from Node | |||
---|---|---|---|
_train_seq List of tuples: [(training-phase1, stop-training-phase1), (training-phase2, stop_training-phase2), ... |
|||
dtype dtype |
|||
input_dim Input dimensions |
|||
output_dim Output dimensions |
|||
supported_dtypes Supported dtypes |
|
Keyword Arguments: k -- number of nearest neighbors to use; the node will raise an MDPWarning if k is smaller than k >= 1 + output_dim + output_dim*(output_dim+1)/2, because in this case a less efficient computation must be used, and the ablgorithm can become unstable r -- regularization constant; as opposed to LLENode, it is not possible to compute this constant automatically; it is only used during execution svd -- if True, use SVD to compute the projection matrix; SVD is slower but more stable verbose -- if True, displays information about the progress of the algorithm output_dim -- number of dimensions to output or a float between 0.0 and 1.0. In the latter case, output_dim specifies the desired fraction of variance to be exaplained, and the final number of output dimensions is known at the end of training (e.g., for 'output_dim=0.95' the algorithm will keep as many dimensions as necessary in order to explain 95% of the input variance)
|
Transform the data list to an array object and reshape it.
|
Home | Trees | Indices | Help |
|
---|
Generated by Epydoc 3.0.1 on Fri Oct 9 06:08:49 2009 | http://epydoc.sourceforge.net |