<?xml version="1.0" encoding="utf-8" ?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
<channel>
<title>kamiさんのブログ</title>
<link>https://ameblo.jp/kami236/</link>
<atom:link href="https://rssblog.ameba.jp/kami236/rss20.xml" rel="self" type="application/rss+xml" />
<atom:link rel="hub" href="http://pubsubhubbub.appspot.com" />
<description>ブログの説明を入力します。</description>
<language>ja</language>
<item>
<title>Overcoming catastrophic forgetting in neural net</title>
<description>
<![CDATA[ <p>Overcoming catastrophic forgetting in neural networks</p><p>I read this paper today.</p><p>&nbsp;</p><p>I think AI should have this capbility:</p><p>1. enlarge its neuron when learing and memoring new tasks.</p><p>&nbsp;</p>
]]>
</description>
<link>https://ameblo.jp/kami236/entry-12419645742.html</link>
<pubDate>Sat, 17 Nov 2018 13:12:52 +0900</pubDate>
</item>
<item>
<title>AI needs memory</title>
<description>
<![CDATA[ <p>&nbsp;</p><p><a href="https://medium.com/@jrodthoughts/memory-in-deep-learning-systems-part-i-the-neuroscience-perspective-aa47fd56ad3c" target="_blank">https://medium.com/@jrodthoughts/memory-in-deep-learning-systems-part-i-the-neuroscience-perspective-aa47fd56ad3c</a></p><p>&nbsp;</p><p><a href="https://towardsdatascience.com/understanding-memory-in-deep-learning-systems-the-neuroscience-psychology-and-technology-78922ae6a1dd" target="_blank">https://towardsdatascience.com/understanding-memory-in-deep-learning-systems-the-neuroscience-psychology-and-technology-78922ae6a1dd(extension version of the first link)</a></p><p>&nbsp;</p><p>&nbsp;</p><p>keywords</p><p>prime memroy, neural turing machine (NTM proposed in 2014 by deepmind with a memory)</p>
]]>
</description>
<link>https://ameblo.jp/kami236/entry-12409293735.html</link>
<pubDate>Wed, 03 Oct 2018 11:38:41 +0900</pubDate>
</item>
</channel>
</rss>
