1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
|
package de.lmu.ifi.dbs.elki.utilities.scaling.outlier;
/*
This file is part of ELKI:
Environment for Developing KDD-Applications Supported by Index-Structures
Copyright (C) 2011
Ludwig-Maximilians-Universität München
Lehr- und Forschungseinheit für Datenbanksysteme
ELKI Development Team
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU Affero General Public License as published by
the Free Software Foundation, either version 3 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 Affero General Public License for more details.
You should have received a copy of the GNU Affero General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
import de.lmu.ifi.dbs.elki.database.ids.DBID;
import de.lmu.ifi.dbs.elki.math.MathUtil;
import de.lmu.ifi.dbs.elki.math.MeanVariance;
import de.lmu.ifi.dbs.elki.result.outlier.OutlierResult;
import de.lmu.ifi.dbs.elki.utilities.documentation.Reference;
import de.lmu.ifi.dbs.elki.utilities.optionhandling.AbstractParameterizer;
import de.lmu.ifi.dbs.elki.utilities.optionhandling.OptionID;
import de.lmu.ifi.dbs.elki.utilities.optionhandling.parameterization.Parameterization;
import de.lmu.ifi.dbs.elki.utilities.optionhandling.parameters.DoubleParameter;
/**
* Scaling that can map arbitrary values to a probability in the range of [0:1].
*
* Transformation is done using the formula max(0, erf(lambda * (x - mean) /
* (stddev * sqrt(2))))
*
* Where mean can be fixed to a given value, and stddev is then computed against
* this mean.
*
* @author Erich Schubert
*/
@Reference(authors="H.-P. Kriegel, P. Kröger, E. Schubert, A. Zimek", title="Interpreting and Unifying Outlier Scores", booktitle="Proc. 11th SIAM International Conference on Data Mining (SDM), Mesa, AZ, 2011", url="http://www.dbs.ifi.lmu.de/~zimek/publications/SDM2011/SDM11-outlier-preprint.pdf")
public class StandardDeviationScaling implements OutlierScalingFunction {
/**
* Parameter to specify a fixed mean to use.
* <p>
* Key: {@code -stddevscale.mean}
* </p>
*/
public static final OptionID MEAN_ID = OptionID.getOrCreateOptionID("stddevscale.mean", "Fixed mean to use in standard deviation scaling.");
/**
* Parameter to specify the lambda value
* <p>
* Key: {@code -stddevscale.lambda}
* </p>
*/
public static final OptionID LAMBDA_ID = OptionID.getOrCreateOptionID("stddevscale.lambda", "Significance level to use for error function.");
/**
* Field storing the fixed mean to use
*/
protected Double fixedmean = null;
/**
* Field storing the lambda value
*/
protected Double lambda = null;
/**
* Mean to use
*/
double mean;
/**
* Scaling factor to use (usually: Lambda * Stddev * Sqrt(2))
*/
double factor;
/**
* Constructor.
*
* @param fixedmean
* @param lambda
*/
public StandardDeviationScaling(Double fixedmean, Double lambda) {
super();
this.fixedmean = fixedmean;
this.lambda = lambda;
}
/**
* Constructor.
*/
public StandardDeviationScaling() {
this(null, 1.0);
}
@Override
public double getScaled(double value) {
assert (factor != 0) : "prepare() was not run prior to using the scaling function.";
if(value <= mean) {
return 0;
}
return Math.max(0, MathUtil.erf((value - mean) / factor));
}
@Override
public void prepare(OutlierResult or) {
if(fixedmean == null) {
MeanVariance mv = new MeanVariance();
for(DBID id : or.getScores().iterDBIDs()) {
double val = or.getScores().get(id);
if(!Double.isNaN(val) && !Double.isInfinite(val)) {
mv.put(val);
}
}
mean = mv.getMean();
factor = lambda * mv.getSampleStddev() * MathUtil.SQRT2;
}
else {
mean = fixedmean;
double sqsum = 0;
int cnt = 0;
for(DBID id : or.getScores().iterDBIDs()) {
double val = or.getScores().get(id);
if(!Double.isNaN(val) && !Double.isInfinite(val)) {
sqsum += (val - mean) * (val - mean);
cnt += 1;
}
}
factor = lambda * Math.sqrt(sqsum / cnt) * MathUtil.SQRT2;
}
}
@Override
public double getMin() {
return 0.0;
}
@Override
public double getMax() {
return 1.0;
}
/**
* Parameterization class.
*
* @author Erich Schubert
*
* @apiviz.exclude
*/
public static class Parameterizer extends AbstractParameterizer {
protected Double fixedmean = null;
protected Double lambda = null;
@Override
protected void makeOptions(Parameterization config) {
super.makeOptions(config);
DoubleParameter meanP = new DoubleParameter(MEAN_ID, true);
if(config.grab(meanP)) {
fixedmean = meanP.getValue();
}
DoubleParameter lambdaP = new DoubleParameter(LAMBDA_ID, 3.0);
if(config.grab(lambdaP)) {
lambda = lambdaP.getValue();
}
}
@Override
protected StandardDeviationScaling makeInstance() {
return new StandardDeviationScaling(fixedmean, lambda);
}
}
}
|
