infomax_ica.h

Go to the documentation of this file.

/* MLPACK 0.2
 *
 * Copyright (c) 2008, 2009 Alexander Gray,
 *                          Garry Boyer,
 *                          Ryan Riegel,
 *                          Nikolaos Vasiloglou,
 *                          Dongryeol Lee,
 *                          Chip Mappus, 
 *                          Nishant Mehta,
 *                          Hua Ouyang,
 *                          Parikshit Ram,
 *                          Long Tran,
 *                          Wee Chin Wong
 *
 * Copyright (c) 2008, 2009 Georgia Institute of Technology
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License as
 * published by the Free Software Foundation; either version 2 of the
 * License, or (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful, but
 * WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
 * 02110-1301, USA.
 */
#ifndef U_INFOMAX_ICA
#define U_INFOMAX_ICA

//#include <math.h>
//#include <limits>
//#include <values.h>
#include <cmath>
#include <climits>
#include "fastlib/fastlib.h"

class TestInfomaxICA; // forward reference

const fx_entry_doc infomax_ica_entries[] = {
  {"lambda", FX_PARAM, FX_DOUBLE, NULL,
   "  Learning rate for infomax method.\n"},
  {"B", FX_PARAM, FX_INT, NULL,
   "  Infomax data window size.\n"},
  {"epsilon", FX_PARAM, FX_DOUBLE, NULL,
   "  Infomax algorithm stop threshold.\n"},
  FX_ENTRY_DOC_DONE
};

const fx_module_doc infomax_ica_doc = {
  infomax_ica_entries, NULL,
  "Performs ICA decomposition using Infomax method.\n"
};

class InfomaxICA {

  friend class TestInfomaxICA;
  
 public:
  InfomaxICA();
  InfomaxICA(double lambda, int B, double epsilon);
  void applyICA(const Matrix &dataset);
  void evaluateICA();
  void displayMatrix(const Matrix &m);
  void displayVector(const Vector &m);
  void getUnmixing(Matrix &w);
  void getSources(const Matrix &dataset, Matrix &s);
  void setLambda(const double lambda);
  void setB(const int b);
  void setEpsilon(const double epsilon);

 private:
  Matrix w_;
  Matrix data_;
  // learning rate 
  double lambda_;
  // block size
  int b_;
  // epsilon for convergence
  double epsilon_;
  // utility functions
  void expM(Matrix &m);
  void addOne(Matrix &m);
  void invertVals(Matrix &m);
  void vectorize(const Matrix &m, Vector &v);
  Matrix eye(index_t dim,double diagVal);
  Matrix sqrtm(const Matrix &m);
  void sphere(Matrix &m);
  Matrix subMeans(const Matrix &m);
  Vector rowMean(const Matrix &m);
  Matrix sampleCovariance(const Matrix &m);
  double w_delta(const Matrix &w_prev, const Matrix &w_pres);
};

#endif