|
1
|
|
|
<?php |
|
2
|
|
|
|
|
3
|
|
|
declare(strict_types=1); |
|
4
|
|
|
|
|
5
|
|
|
namespace Phpml\Classification; |
|
6
|
|
|
|
|
7
|
|
|
use Phpml\Helper\Predictable; |
|
8
|
|
|
use Phpml\Helper\Trainable; |
|
9
|
|
|
use Phpml\Math\Distance; |
|
10
|
|
|
use Phpml\Math\Distance\Euclidean; |
|
11
|
|
|
|
|
12
|
|
|
class KNearestNeighbors implements Classifier |
|
13
|
|
|
{ |
|
14
|
|
|
use Trainable, Predictable; |
|
15
|
|
|
|
|
16
|
|
|
/** |
|
17
|
|
|
* @var int |
|
18
|
|
|
*/ |
|
19
|
|
|
private $k; |
|
20
|
|
|
|
|
21
|
|
|
/** |
|
22
|
|
|
* @var Distance |
|
23
|
|
|
*/ |
|
24
|
|
|
private $distanceMetric; |
|
25
|
|
|
|
|
26
|
|
|
/** |
|
27
|
|
|
* @param int $k |
|
28
|
|
|
* @param Distance|null $distanceMetric (if null then Euclidean distance as default) |
|
29
|
|
|
*/ |
|
30
|
|
|
public function __construct(int $k = 3, Distance $distanceMetric = null) |
|
31
|
|
|
{ |
|
32
|
|
|
if (null === $distanceMetric) { |
|
33
|
|
|
$distanceMetric = new Euclidean(); |
|
34
|
|
|
} |
|
35
|
|
|
|
|
36
|
|
|
$this->k = $k; |
|
37
|
|
|
$this->samples = []; |
|
38
|
|
|
$this->targets = []; |
|
39
|
|
|
$this->distanceMetric = $distanceMetric; |
|
40
|
|
|
} |
|
41
|
|
|
|
|
42
|
|
|
/** |
|
43
|
|
|
* @param array $samples |
|
44
|
|
|
* @param array $targets |
|
45
|
|
|
*/ |
|
46
|
|
|
public function train(array $samples, array $targets) |
|
47
|
|
|
{ |
|
48
|
|
|
$this->samples = array_merge($this->samples, $samples); |
|
49
|
|
|
$this->targets = array_merge($this->targets, $targets); |
|
50
|
|
|
} |
|
51
|
|
|
|
|
52
|
|
|
/** |
|
53
|
|
|
* @param array $sample |
|
54
|
|
|
* |
|
55
|
|
|
* @return mixed |
|
56
|
|
|
*/ |
|
57
|
|
|
protected function predictSample(array $sample) |
|
58
|
|
|
{ |
|
59
|
|
|
$distances = $this->kNeighborsDistances($sample); |
|
60
|
|
|
|
|
61
|
|
|
$predictions = array_combine(array_values($this->targets), array_fill(0, count($this->targets), 0)); |
|
62
|
|
|
|
|
63
|
|
|
foreach ($distances as $index => $distance) { |
|
64
|
|
|
++$predictions[$this->targets[$index]]; |
|
65
|
|
|
} |
|
66
|
|
|
|
|
67
|
|
|
arsort($predictions); |
|
68
|
|
|
reset($predictions); |
|
69
|
|
|
|
|
70
|
|
|
return key($predictions); |
|
71
|
|
|
} |
|
72
|
|
|
|
|
73
|
|
|
/** |
|
74
|
|
|
* @param array $sample |
|
75
|
|
|
* |
|
76
|
|
|
* @return array |
|
77
|
|
|
* |
|
78
|
|
|
* @throws \Phpml\Exception\InvalidArgumentException |
|
79
|
|
|
*/ |
|
80
|
|
|
private function kNeighborsDistances(array $sample) |
|
81
|
|
|
{ |
|
82
|
|
|
$distances = []; |
|
83
|
|
|
|
|
84
|
|
|
foreach ($this->samples as $index => $neighbor) { |
|
85
|
|
|
$distances[$index] = $this->distanceMetric->distance($sample, $neighbor); |
|
86
|
|
|
} |
|
87
|
|
|
|
|
88
|
|
|
asort($distances); |
|
89
|
|
|
|
|
90
|
|
|
return array_slice($distances, 0, $this->k, true); |
|
91
|
|
|
} |
|
92
|
|
|
} |
|
93
|
|
|
|
php-ai /
php-ml
