|
所在平台: Udemy |
课程主页: https://www.udemy.com/course/linear-algebra-data-science/
课程评论：没有评论
课程名称：数学 0-1：数据科学与机器学习的线性代数 课程概述： 在进入机器学习和数据科学的过程中，许多人会发现数学知识的要求非常高。无论是从未学习过相关数学，还是早已遗忘，这都是一个常见的情境。为了帮助大家解决这个问题，我开发了这门课程。线性代数是机器学习最重要的数学基础之一，它为理解概率和统计提供了必要的支撑，这也是数据科学的基础。在数据科学中，“数据”通过矩阵和向量来表示，而这些正是本课程的核心研究对象。如果你希望在机器学习中深入发展，而不仅仅是复制博客和教程中的库代码，掌握线性代数是必不可少的。 通常，在STEM学院的课程中，线性代数会分为多个学期进行学习。而在这门课程中，我提炼出最必要的知识，使你可以在短短几个小时内掌握所有必备的内容。这门课程将涵盖线性方程组、矩阵运算（点积、逆矩阵、转置、行列式、迹）、低秩近似、正定性与负定性、特征值与特征向量等。课程还将包括一些通常在常规大学课程中不涉及的与机器学习相关的材料，例如这些概念如何应用于GPT-4，以及如何使用LoRA微调现代神经网络（如扩散模型和大型语言模型）。我们还会使用Python编程语言演示课程中的许多概念（无需掌握Python基础）。 总之，这门课程不同于传统的线性代数课程，它只专注于最实用、最具影响力的主题，提供与机器学习和数据科学直接相关的技能，帮助你能够立即应用这些知识。准备好了吗？让我们开始吧！ 建议的先备知识：扎实的高中数学基础（函数、代数、三角函数）
Common scenario: You try to get into machine learning and data science, but there's SO MUCH MATH.Either you never studied this math, or you studied it so long ago you've forgotten it all.What do you do?Well my friends, that is why I created this course.Linear Algebra is one of the most important math prerequisites for machine learning. It's required to understand probability and statistics, which form the foundation of data science.The "data" in data science is represented using matrices and vectors, which are the central objects of study in this course.If you want to do machine learning beyond just copying library code from blogs and tutorials, you must know linear algebra.In a normal STEM college program, linear algebra is split into multiple semester-long courses.Luckily, I've refined these teachings into just the essentials, so that you can learn everything you need to know on the scale of hours instead of semesters.This course will cover systems of linear equations, matrix operations (dot product, inverse, transpose, determinant, trace), low-rank approximations, positive-definiteness and negative-definiteness, and eigenvalues and eigenvectors. It will even include machine learning-focused material you wouldn't normally see in a regular college course, such as how these concepts apply to GPT-4, and fine-tuning modern neural networks like diffusion models (for generative AI art) and LLMs (Large Language Models) using LoRA. We will even demonstrate many of the concepts in this course using the Python programming language (don't worry, you don't need to know Python for this course). In other words, instead of the dry old college version of linear algebra, this course takes just the most practical and impactful topics, and provides you with skills directly applicable to machine learning and data science, so you can start applying them today.Are you ready?Let's go!Suggested prerequisites:Firm understanding of high school math (functions, algebra, trigonometry)
