<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Matrices on Rik Kisnah - Blog</title><link>https://www.rik-kisnah.ai/tags/matrices/</link><description>Recent content in Matrices on Rik Kisnah - Blog</description><generator>Hugo</generator><language>en</language><lastBuildDate>Sat, 01 May 1999 00:00:00 +0000</lastBuildDate><atom:link href="https://www.rik-kisnah.ai/tags/matrices/feed.xml" rel="self" type="application/rss+xml"/><item><title>Matrix Neural Net Dreams: Reflections on Early AI</title><link>https://www.rik-kisnah.ai/posts/matrix-neural-net-dreams/</link><pubDate>Sat, 01 May 1999 00:00:00 +0000</pubDate><guid>https://www.rik-kisnah.ai/posts/matrix-neural-net-dreams/</guid><description>Machine Learning Before It Was Cool (1999) While everyone worried about Y2K, I was playing with neural networks at NTU. Not because I thought they&amp;rsquo;d ship in products—they were academic curiosities then. But because the math was beautiful.
The core insight: if you could represent a problem as matrices, you could teach a network to solve it via backpropagation. Weights → matrix multiply → gradient descent → improved weights. Repeat until convergence.</description></item></channel></rss>