001/** 002 * Copyright (c) 2011, The University of Southampton and the individual contributors. 003 * All rights reserved. 004 * 005 * Redistribution and use in source and binary forms, with or without modification, 006 * are permitted provided that the following conditions are met: 007 * 008 * * Redistributions of source code must retain the above copyright notice, 009 * this list of conditions and the following disclaimer. 010 * 011 * * Redistributions in binary form must reproduce the above copyright notice, 012 * this list of conditions and the following disclaimer in the documentation 013 * and/or other materials provided with the distribution. 014 * 015 * * Neither the name of the University of Southampton nor the names of its 016 * contributors may be used to endorse or promote products derived from this 017 * software without specific prior written permission. 018 * 019 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND 020 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED 021 * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE 022 * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR 023 * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES 024 * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; 025 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON 026 * ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT 027 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS 028 * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 029 */ 030package org.openimaj.ml.linear.learner.loss; 031 032import gov.sandia.cognition.math.matrix.Matrix; 033 034import org.apache.log4j.Logger; 035 036public class SquareMissingLossFunction extends LossFunction { 037 Logger logger = Logger.getLogger(SquareMissingLossFunction.class); 038 039 @Override 040 public Matrix gradient(Matrix W) { 041 final Matrix resid = X.times(W).minus(Y); 042 if (this.bias != null) 043 resid.plusEquals(this.bias); 044 for (int r = 0; r < Y.getNumRows(); r++) { 045 final double yc = Y.getElement(r, 0); 046 if (Double.isNaN(yc)) { 047 resid.setElement(r, 0, 0); 048 } 049 } 050 return X.transpose().times(resid); 051 } 052 053 @Override 054 public double eval(Matrix W) { 055 Matrix v; 056 if (W == null) { 057 v = this.X; 058 } 059 else { 060 v = X.times(W); 061 } 062 final Matrix vWithoutBias = v.clone(); 063 if (this.bias != null) 064 v.plusEquals(this.bias); 065 double sum = 0; 066 for (int r = 0; r < Y.getNumRows(); r++) { 067 for (int c = 0; c < Y.getNumColumns(); c++) { 068 final double yr = Y.getElement(r, c); 069 if (!Double.isNaN(yr)) { 070 final double val = v.getElement(r, c); 071 final double valNoBias = vWithoutBias.getElement(r, c); 072 final double delta = yr - val; 073 logger.debug( 074 String.format( 075 "yr=%d,y=%3.2f,v=%3.2f,v(no bias)=%2.5f,delta=%2.5f", 076 r, yr, val, valNoBias, delta 077 ) 078 ); 079 sum += delta * delta; 080 } 081 } 082 } 083 return sum; 084 } 085 086 @Override 087 public boolean isMatrixLoss() { 088 return false; 089 } 090 091}