mirror of
https://github.com/fofolee/uTools-Manuals.git
synced 2025-06-19 14:22:43 +08:00
77 lines
4.0 KiB
HTML
77 lines
4.0 KiB
HTML
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
|
|
<html>
|
|
<head>
|
|
<meta http-equiv="content-type" content="text/html; charset=UTF-8">
|
|
<title>返回级联候选变化分数</title>
|
|
</head>
|
|
<body class="docs"><div id="layout">
|
|
<div id="layout-content"><div id="function.fann-get-cascade-candidate-change-fraction" class="refentry">
|
|
<div class="refnamediv">
|
|
<h1 class="refname">fann_get_cascade_candidate_change_fraction</h1>
|
|
<p class="verinfo">(PECL fann >= 1.0.0)</p><p class="refpurpose"><span class="refname">fann_get_cascade_candidate_change_fraction</span> — <span class="dc-title">返回级联候选变化分数</span></p>
|
|
|
|
</div>
|
|
|
|
<div class="refsect1 description" id="refsect1-function.fann-get-cascade-candidate-change-fraction-description">
|
|
<h3 class="title">说明</h3>
|
|
<div class="methodsynopsis dc-description">
|
|
<span class="methodname"><strong>fann_get_cascade_candidate_change_fraction</strong></span>
|
|
( <span class="methodparam"><span class="type">resource</span> <code class="parameter">$ann</code></span>
|
|
) : <span class="type">float</span></div>
|
|
|
|
<p class="para rdfs-comment">
|
|
返回级联候选变化分数是一个介于0到1之间的数字。该数字决定了在训练候选神经元使用 <span class="function"><a href="fann_get_cascade_candidate_stagnation_epochs.html" class="function">fann_get_cascade_candidate_stagnation_epochs()</a></span> 函数时 <span class="function"><a href="fann_get_mse.html" class="function">fann_get_MSE()</a></span> 将会改变多大的分数,是为了不让这个训练停止。如果训练停止，候选神经元的训练将会被终止并且会选出最好的候选。
|
|
</p>
|
|
<p class="para">
|
|
这意味在<span class="function"><a href="fann_get_cascade_candidate_stagnation_epochs.html" class="function">fann_get_cascade_candidate_stagnation_epochs()</a></span>期间如果 MSE 被<span class="function"><strong>fann_get_cascade_candidate_change_fraction()</strong></span>的分数更改，候选神经元的训练会因为训练的停滞而被停止。
|
|
</p>
|
|
<p class="para">
|
|
如果级联候选变化分数很低，这个候选神经元将训练越多，如果这个分数很高则训练的越少。
|
|
</p>
|
|
<p class="para">
|
|
默认的级联候选变化分数是 0.01, 等于 MSE 中1%的变化值。
|
|
</p>
|
|
</div>
|
|
|
|
|
|
<div class="refsect1 parameters" id="refsect1-function.fann-get-cascade-candidate-change-fraction-parameters">
|
|
<h3 class="title">参数</h3>
|
|
<dl>
|
|
|
|
|
|
<dt>
|
|
<code class="parameter">ann</code></dt>
|
|
|
|
<dd>
|
|
|
|
<p class="para">Neural network <span class="type"><a href="language.types.resource.html" class="type resource">resource</a></span>.</p>
|
|
</dd>
|
|
|
|
|
|
</dl>
|
|
|
|
</div>
|
|
|
|
|
|
<div class="refsect1 returnvalues" id="refsect1-function.fann-get-cascade-candidate-change-fraction-returnvalues">
|
|
<h3 class="title">返回值</h3>
|
|
<p class="para">
|
|
成功，则返回级联候选变化分数，错误则返回 <strong><code>FALSE</code></strong> .
|
|
</p>
|
|
</div>
|
|
|
|
|
|
<div class="refsect1 seealso" id="refsect1-function.fann-get-cascade-candidate-change-fraction-seealso">
|
|
<h3 class="title">参见</h3>
|
|
<p class="para">
|
|
<ul class="simplelist">
|
|
<li class="member"><span class="function"><a href="fann_set_cascade_candidate_change_fraction.html" class="function" rel="rdfs-seeAlso">fann_set_cascade_candidate_change_fraction()</a> - 设置级联候选更改分数</span></li>
|
|
<li class="member"><span class="function"><a href="fann_get_mse.html" class="function" rel="rdfs-seeAlso">fann_get_MSE()</a> - 从网络中读取均方误差。</span></li>
|
|
<li class="member"><span class="function"><a href="fann_get_cascade_candidate_stagnation_epochs.html" class="function" rel="rdfs-seeAlso">fann_get_cascade_candidate_stagnation_epochs()</a> - 返回层叠候选停滞周期的数量</span></li>
|
|
</ul>
|
|
</p>
|
|
</div>
|
|
|
|
|
|
|
|
</div></div></div></body></html> |