<div data-page-id="doxcnlNXBd0w44WNLEwJZE33Xgc" data-docx-has-block-data="true"> <div class="image-uploaded gallery old-record-id-doxcnOe4aAIskkmuaExAhOXBbOc" style="text-align: justify;" data-type="image" data-ace-gallery-json="{&quot;items&quot;:[{&quot;uuid&quot;:&quot;f35c5b54-f17f-48b0-b95d-585593b5b563&quot;,&quot;height&quot;:618,&quot;width&quot;:778,&quot;currHeight&quot;:618,&quot;currWidth&quot;:778,&quot;natrualHeight&quot;:618,&quot;natrualWidth&quot;:778,&quot;pluginName&quot;:&quot;imageUpload&quot;,&quot;scale&quot;:1,&quot;src&quot;:&quot;https%3A%2F%2Finternal-api-drive-stream.feishu.cn%2Fspace%2Fapi%2Fbox%2Fstream%2Fdownload%2Fall%2FboxcnEYv41Bph6StiEIcLngRTzg%2F%3Fmount_node_token%3DdoxcnOe4aAIskkmuaExAhOXBbOc%26mount_point%3Ddocx_image&quot;,&quot;file_token&quot;:&quot;boxcnEYv41Bph6StiEIcLngRTzg&quot;,&quot;image_type&quot;:&quot;image/png&quot;,&quot;size&quot;:947787,&quot;comments&quot;:[]}]}"><span class="p"><img src="https://info.compassedu.hk/sucai/content/1658819350683/1658819350683.png" width="778" height="618" /></span></div> <div class="ace-line ace-line old-record-id-doxcnuUuoI4Agi0MeQGq1wjTgJD" style="text-align: justify;"><span class="p">&nbsp;</span></div> <div class="ace-line ace-line old-record-id-doxcng0cGsqCWQyOGMXXVoLIgYd" style="text-align: justify;"><span class="p">人類和機器智能共同努力發(fā)現(xiàn)了40，000個環(huán)形星系，科學家們今天將在全國天文學會議上宣布。</span></div> <div class="ace-line ace-line old-record-id-doxcn2WOYMa0UogUow7ckhdDITc" style="text-align: justify;"><span class="p">&nbsp;</span></div> <div class="ace-line ace-line old-record-id-doxcn2scS66ia0ai247VauskYXf" style="text-align: justify;"><span class="p">曼徹斯特大學的Mike Walmsley博士將介紹這項新工作，描述這種&ldquo;半機械人&rdquo;方法如何測量數(shù)百萬個星系的形狀。</span></div> <div class="ace-line ace-line old-record-id-doxcncQUEGoCgUmEqiS4tHIPKuc" style="text-align: justify;"><span class="p">&nbsp;</span></div> <div class="ace-line ace-line old-record-id-doxcnUoWaKUkouWakmOfYKDR2xh" style="text-align: justify;"><span class="p">星系過著混亂的生活。與其他星系的碰撞以及來自超大質量黑洞的能量爆發(fā)破壞了數(shù)十億顆恒星的顏色和軌道，留下了志愿者在銀河動物園網(wǎng)站上搜索的標記。但是，確切地了解哪些宇宙事件導致哪些標記需要數(shù)百萬張測量圖像 - 比人類所能搜索的還要多。</span></div> <div class="ace-line ace-line old-record-id-doxcnKgEkCw6SyyIuYhY2PmVnBc" style="text-align: justify;"><span class="p">&nbsp;</span></div> <div class="ace-line ace-line old-record-id-doxcnwQsSqQ4m6W08Y6i21LlAbb" style="text-align: justify;"><span class="p">為了提供幫助，沃爾姆斯利博士呼吁&ldquo;銀河動物園&rdquo;，該項目招募了數(shù)十萬志愿者來測量數(shù)百萬個星系的形態(tài)（形狀）。</span></div> <div class="ace-line ace-line old-record-id-doxcnUCii6gSY6m2ucLh9a3gUpc" style="text-align: justify;"><span class="p">&nbsp;</span></div> <blockquote class=" old-record-id-doxcnaMYGEWyYm0WM0ajZV2m32g" data-type="quote_container"> <div class="ace-line ace-line old-record-id-doxcn24Ic20UocwSg2RAXIIAb1c" style="text-align: center;"><span class="h6"><em>通過Zoobot，Galaxy Zoo志愿者將幫助其他天文學家解決我們從未想過要問的問題。</em></span></div> <div class="ace-line ace-line old-record-id-doxcnuyocSU4CIIMeSimjEVHILc" style="text-align: center;"><em><strong><span class="h6">&mdash;&mdash;邁克&middot;沃姆斯利博士</span></strong></em></div> </blockquote> <div class="ace-line ace-line old-record-id-doxcnsK6IEIAOoGC4I5y49Moxnf" style="text-align: justify;"><span class="p">&nbsp;</span></div> <div class="ace-line ace-line old-record-id-doxcnm2yAaWEEwAoSWymVKdnD2b" style="text-align: justify;"><span class="p">十年的Galaxy Zoo志愿者測量（總計超過9600萬次點擊）被用來創(chuàng)建一個自動助手 - 一種新的AI算法。該算法被親切地命名為&ldquo;Zoobot&rdquo;，不僅可以準確地預測志愿者會說什么，還可以理解它可能出錯的地方。</span></div> <div class="ace-line ace-line old-record-id-doxcnsW8ma4sKMOc0BcbKvCABse" style="text-align: justify;"><span class="p">&nbsp;</span></div> <div class="ace-line ace-line old-record-id-doxcnKgMqssg446eQMdYupqhWpe" style="text-align: justify;"><span class="p">40，000個罕見的環(huán)形星系的發(fā)現(xiàn)是以前已知的六倍。環(huán)需要數(shù)十億年才能形成，并在星系 - 星系碰撞中被摧毀，因此這個巨大的新樣本將有助于揭示孤立的星系是如何演化的。該數(shù)據(jù)集還將告訴科學家星系如何更普遍地老化。</span></div> <div class="ace-line ace-line old-record-id-doxcnwgymsUiiWsyUmKroQQTuEz" style="text-align: justify;"><span class="p">&nbsp;</span></div> <div class="ace-line ace-line old-record-id-doxcnm2WsUQCcSWOyk3nGUL59Gf" style="text-align: justify;"><span class="p">Zoobot旨在一次又一次地為新的科學目標進行再培訓。就像音樂家可以比他們的第一個樂器更快地學習新樂器一樣，Zoobot可以輕松學習回答新的形狀問題，因為它已經(jīng)學會了回答50多個不同的問題。Walmsley博士說：&ldquo;通過Zoobot，Galaxy Zoo志愿者將幫助其他天文學家解決我們從未想過要問的問題。</span></div> </div>