source: trunk/doc/concepts.doxygen @ 1497

Last change on this file since 1497 was 1487, checked in by Jari Häkkinen, 14 years ago

Addresses #436. GPL license copy reference should also be updated.

  • Property svn:eol-style set to native
  • Property svn:keywords set to Id
File size: 4.0 KB
Line 
1// $Id: concepts.doxygen 1487 2008-09-10 08:41:36Z jari $
2//
3// Copyright (C) 2008 Peter Johansson, Markus Ringnér
4//
5// This file is part of the yat library, http://dev.thep.lu.se/yat
6//
7// The yat library is free software; you can redistribute it and/or
8// modify it under the terms of the GNU General Public License as
9// published by the Free Software Foundation; either version 3 of the
10// License, or (at your option) any later version.
11//
12// The yat library is distributed in the hope that it will be useful,
13// but WITHOUT ANY WARRANTY; without even the implied warranty of
14// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
15// General Public License for more details.
16//
17// You should have received a copy of the GNU General Public License
18// along with yat. If not, see <http://www.gnu.org/licenses/>.
19
20
21/**
22\page Concepts Concepts
23
24This page lists all the C++ concepts in the yat project.
25
26- \subpage concept_distance
27- \subpage concept_neighbor_weighting
28*/
29
30
31/**
32\page concept_distance Distance
33
34\section Description
35
36\ref concept_distance is a concept for classes implementing different
37alternatives to calculate the distance between two points.
38
39\section Requirements
40
41Classes modelling the concept \ref concept_distance should implement
42the following public function:
43
44\verbatim
45template<typename Iter1, typename Iter2>
46double  operator() (Iter1 beg1, Iter1 end1, Iter2 beg2) const
47\endverbatim
48
49This function should calculate and return the distance between
50elements of two ranges. The first range is given by [\a beg1, \a end1)
51and the second range starts with \a beg2 and has the same length as
52the first range. The function should support iterators of the category
53std::forward_iterator. The function should provide both a fast
54calculation for unweighted iterators and a calculation for weighted
55iterators. The latter correspond to the case where elements in a range
56have both a value and a weight. The selection between unweighted and
57weighted implementations should utilize
58theplu::yat::utility::unweighted_iterator_tag and
59theplu::yat::utility::weighted_iterator_tag. Moreover
60theplu::yat::utility::weighted_if_any2 should be utilized to provide a
61weighted implementation if any of the two ranges is weighted, and an
62unweighted implementation when both ranges are unweighted.
63
64\section Implementations
65
66Examples of classes modelling the concept \ref concept_distance
67include theplu::yat::statistics::PearsonDistance and
68theplu::yat::statistics::EuclideanDistance.
69
70\see \ref weighted_distance
71
72*/
73
74/**
75\page concept_neighbor_weighting Neighbor Weighting Method
76
77\section Description
78
79\ref concept_neighbor_weighting is a concept used in connection with
80theplu::yat::classifier::KNN - classes used as the template argument
81NeighborWeighting should implement this concept.
82
83\section Requirements
84
85Classes modelling the concept \ref concept_neighbor_weighting should
86implement the following public function:
87 
88\verbatim   
89void operator()(const utility::VectorBase& distance, const std::vector<size_t> k_sorted,
90                const Target& target, utility::VectorMutable& prediction) const
91\endverbatim
92
93For a test sample, this function should calculate a total vote
94(i.e. based on all k nearest neighbors) for each class. The vector \a
95distance contains the distances from a test sample to all training
96samples. The vector \a k_sorted contains the indices (for both \a
97distance and \a target) to the k training samples with the smallest
98distances to the test sample. The class for each training sample is
99given by \a target, which is sorted in the same sample order as \a
100distance. For each class the function calculates a total vote based on
101the the nearest neighbors of the test sample that belong to the
102class. The total vote for each class is stored in the vector \a
103prediction.  The function should be implemented such that nearest
104neighbors with distance infinity do not vote.
105
106\section Implementations
107
108Examples of classes modelling the concept \ref
109concept_neighbor_weighting include
110theplu::yat::classifier::KNN_Uniform,
111theplu::yat::classifier::KNN_ReciprocalDistance and
112theplu::yat::classifier::KNN_ReciprocalRank.
113
114*/
Note: See TracBrowser for help on using the repository browser.