<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Machine Learning on Bytes of Life</title><link>https://yangyangli.top/categories/machine-learning/</link><description>Recent content in Machine Learning on Bytes of Life</description><generator>Hugo -- gohugo.io</generator><language>en</language><copyright>2026 &lt;a class='hover:underline hover:decoration-primary-400 hover:text-primary-500' href=https://yangyangli.top target=_blank rel='noopener noreferrer'&gt;Yangyang Li&lt;/a&gt;</copyright><lastBuildDate>Thu, 07 May 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://yangyangli.top/categories/machine-learning/index.xml" rel="self" type="application/rss+xml"/><item><title>DeepChopper: A Genomic Language Model that Cleans Up Nanopore Direct RNA Sequencing</title><link>https://yangyangli.top/posts/027-deepchopper/</link><pubDate>Thu, 07 May 2026 00:00:00 +0000</pubDate><guid>https://yangyangli.top/posts/027-deepchopper/</guid><description>&lt;h2 class="relative group"&gt;The chimera mystery
&lt;div id="the-chimera-mystery" class="anchor"&gt;&lt;/div&gt;
&lt;span
class="absolute top-0 w-6 transition-opacity opacity-0 ltr:-left-6 rtl:-right-6 not-prose group-hover:opacity-100 select-none"&gt;
&lt;a class="group-hover:text-primary-300 dark:group-hover:text-neutral-700 !no-underline" href="#the-chimera-mystery" aria-label="Anchor"&gt;#&lt;/a&gt;
&lt;/span&gt;
&lt;/h2&gt;
&lt;p&gt;Direct RNA sequencing (dRNA-seq) on Oxford Nanopore looks, on paper, like a transcriptomics dream.
You sequence native RNA molecules end to end, you keep the modifications, and you skip every reverse-transcription and PCR step that has been quietly polluting short-read data for years.
For a while, that was the story we were telling ourselves.&lt;/p&gt;</description><media:content xmlns:media="http://search.yahoo.com/mrss/" url="https://yangyangli.top/posts/027-deepchopper/featured.jpg"/></item><item><title>A language model enables accurate structural variant detection in whole-genome amplified long-read sequencing</title><link>https://yangyangli.top/projects/006-chimeralm/</link><pubDate>Thu, 23 Jan 2025 00:00:00 +0000</pubDate><guid>https://yangyangli.top/projects/006-chimeralm/</guid><description/><media:content xmlns:media="http://search.yahoo.com/mrss/" url="https://yangyangli.top/projects/006-chimeralm/featured.png"/></item><item><title>Genomic Language Model Mitigates Chimera Artifacts in Nanopore Direct RNA Sequencing</title><link>https://yangyangli.top/projects/005-deepchopper/</link><pubDate>Thu, 31 Oct 2024 00:00:00 +0000</pubDate><guid>https://yangyangli.top/projects/005-deepchopper/</guid><description/><media:content xmlns:media="http://search.yahoo.com/mrss/" url="https://yangyangli.top/projects/005-deepchopper/featured.png"/></item><item><title>CUDA for Deep Learning Inference in Rust and C++</title><link>https://yangyangli.top/posts/022-cuda-configuration-for-rust-and-cpp/</link><pubDate>Tue, 29 Aug 2023 00:00:00 +0000</pubDate><guid>https://yangyangli.top/posts/022-cuda-configuration-for-rust-and-cpp/</guid><description>&lt;h2 class="relative group"&gt;1. Deep Learning Inference
&lt;div id="1-deep-learning-inference" class="anchor"&gt;&lt;/div&gt;
&lt;span
class="absolute top-0 w-6 transition-opacity opacity-0 ltr:-left-6 rtl:-right-6 not-prose group-hover:opacity-100 select-none"&gt;
&lt;a class="group-hover:text-primary-300 dark:group-hover:text-neutral-700 !no-underline" href="#1-deep-learning-inference" aria-label="Anchor"&gt;#&lt;/a&gt;
&lt;/span&gt;
&lt;/h2&gt;
&lt;p&gt;Currently, both Rust and C++ are emerging as noteworthy contenders in the realm
of deep learning, primarily due to their efficiency despite Python&amp;rsquo;s prevailing
dominance in model training.
While Python continues to commandeer the training phase, it lags in performance
during inference.
Large Language Models (LLMs), such as ChatGPT, have burgeoned since their
inception, instigating a competitive frenzy among corporations and research
organizations alike.
This has led to an explosion of various LLMs, although not all exhibit equal
utility or value.&lt;/p&gt;</description><media:content xmlns:media="http://search.yahoo.com/mrss/" url="https://yangyangli.top/posts/022-cuda-configuration-for-rust-and-cpp/featured.jpg"/></item><item><title>PCA by Python</title><link>https://yangyangli.top/posts/010-pca-by-python/</link><pubDate>Sun, 04 Apr 2021 00:00:00 +0000</pubDate><guid>https://yangyangli.top/posts/010-pca-by-python/</guid><description>&lt;h2 class="relative group"&gt;1. Introduction
&lt;div id="1-introduction" class="anchor"&gt;&lt;/div&gt;
&lt;span
class="absolute top-0 w-6 transition-opacity opacity-0 ltr:-left-6 rtl:-right-6 not-prose group-hover:opacity-100 select-none"&gt;
&lt;a class="group-hover:text-primary-300 dark:group-hover:text-neutral-700 !no-underline" href="#1-introduction" aria-label="Anchor"&gt;#&lt;/a&gt;
&lt;/span&gt;
&lt;/h2&gt;
&lt;p&gt;This article records two methods of PCA analysis using Python, and visualizes 2-dimensional results&lt;/p&gt;</description><media:content xmlns:media="http://search.yahoo.com/mrss/" url="https://yangyangli.top/posts/010-pca-by-python/featured.png"/></item><item><title>Face Recognition in Video</title><link>https://yangyangli.top/posts/015-face-recognition-in-video/</link><pubDate>Thu, 20 Jun 2019 00:00:00 +0000</pubDate><guid>https://yangyangli.top/posts/015-face-recognition-in-video/</guid><description>&lt;h2 class="relative group"&gt;1. 准备工作
&lt;div id="1-准备工作" class="anchor"&gt;&lt;/div&gt;
&lt;span
class="absolute top-0 w-6 transition-opacity opacity-0 ltr:-left-6 rtl:-right-6 not-prose group-hover:opacity-100 select-none"&gt;
&lt;a class="group-hover:text-primary-300 dark:group-hover:text-neutral-700 !no-underline" href="#1-%e5%87%86%e5%a4%87%e5%b7%a5%e4%bd%9c" aria-label="Anchor"&gt;#&lt;/a&gt;
&lt;/span&gt;
&lt;/h2&gt;
&lt;p&gt;使用&lt;strong&gt;Python&lt;/strong&gt;实现需求，环境需求比较严苛，所需要的主要依赖：&lt;/p&gt;
&lt;p&gt;&lt;strong&gt;Requirements&lt;/strong&gt;&lt;/p&gt;
&lt;ul&gt;
&lt;li&gt;&lt;strong&gt;dlib&lt;/strong&gt; 由 C++编写，提供了和机器学习、数值计算、图模型算法、图像处理等领域相关的一系列功能&lt;/li&gt;
&lt;li&gt;&lt;strong&gt;face-recognition&lt;/strong&gt; 已经经过训练成熟的识别人脸的工具&lt;/li&gt;
&lt;li&gt;&lt;strong&gt;imutils&lt;/strong&gt; 用来操作系统文件和文件夹&lt;/li&gt;
&lt;li&gt;&lt;strong&gt;Opencv&lt;/strong&gt;是用来操作视频流，并将视频流转换格式&lt;/li&gt;
&lt;/ul&gt;
&lt;h3 class="relative group"&gt;1.1 环境搭建
&lt;div id="11-环境搭建" class="anchor"&gt;&lt;/div&gt;
&lt;span
class="absolute top-0 w-6 transition-opacity opacity-0 ltr:-left-6 rtl:-right-6 not-prose group-hover:opacity-100 select-none"&gt;
&lt;a class="group-hover:text-primary-300 dark:group-hover:text-neutral-700 !no-underline" href="#11-%e7%8e%af%e5%a2%83%e6%90%ad%e5%bb%ba" aria-label="Anchor"&gt;#&lt;/a&gt;
&lt;/span&gt;
&lt;/h3&gt;
&lt;ol&gt;
&lt;li&gt;
&lt;p&gt;通过 pip 进行安装，在使用 pip 进行安装时，应该注意切换到国内源进行下载，提高下载速度，下面分享一下当前国内的 pip 源：&lt;/p&gt;</description><media:content xmlns:media="http://search.yahoo.com/mrss/" url="https://yangyangli.top/posts/015-face-recognition-in-video/featured.png"/></item></channel></rss>