|
所在平台: Udemy |
课程主页: https://www.udemy.com/course/build-neural-networks-from-scratch-with-python-step-by-step/
课程评论：没有评论
**Coursera 课程总结：Python 从零开始构建神经网络** 本课程由拥有 25 年丰富经验的高级软件工程师 Loek van den Ouweland 主讲，旨在教授学员如何使用纯 Python 语言从头开始构建神经网络。与许多号称“从零开始”却依赖外部库的教程不同，本课程强调逐步构建，从一个简单的四行代码神经网络出发，逐步演进为一个能够识别手写数字的人工智能网络。 **核心学习内容：** * **神经网络基础：** 涵盖前馈（Feed forward）、成本函数（Cost functions）、反向传播（Back propagation）、隐藏层（Hidden layers）、线性回归（Linear regression）、梯度下降（Gradient descent）和矩阵乘法（Matrix multiplication）等核心概念。 * **循序渐进的搭建过程：** 课程以极其详细的步骤展示了从基础概念到复杂网络的演变，确保学员理解每一行代码的作用。 * **无框架限制：** 完全不依赖任何第三方神经网络库或框架，让学员深入理解神经网络的内在机制。 **课程特点：** * **零基础友好：** 适合希望深入理解神经网络工作原理的开发者，或者即便使用框架但想了解其底层参数含义的开发者。 * **清晰易懂：** 采用“一步一步”的教学方式，确保初学者也能轻松掌握，避免了许多教程中“代码堆砌”导致的学习障碍。 * **实践导向：** 通过实际的代码构建过程，帮助学员掌握在任何编程语言中实现神经网络的能力。 **课程产出：** 完成本课程后，学员将能够： * 深刻理解反向传播和梯度下降等关键神经网络概念。 * 在没有任何框架或库的帮助下，用任何编程语言构建神经网络。 * 通过调整成本函数和添加隐藏层来优化网络配置。 **目标受众：** * 希望了解神经网络内在机制的开发者。 * 希望避免使用神经网络库和框架的开发者。 * 使用框架但想深入理解其各个参数含义的开发者。 **课程时长：** 3 小时视频内容，无练习。
You will learn how to build Neural Networks with Python. Without the need for any library, you will see how a simple neural network from 4 lines of code, evolves into a artificial intelligence network that is able to recognize handwritten digits.During this process, you will learn concepts like: Feed forward, Cost functions, Back propagation, Hidden layers, Linear regression, Gradient descent and Matrix multiplication. And all this with plain Python.Target audienceDevelopers who especially benefit from this course, are:Developer who want to learn the mechanics of neural networksDevelopers who want to avoid using neural network libraries and frameworksOr developers who use frameworks but want to learn the meaning of the individual network parametersChallengesMany tutorials claim to start from scratch, but import external libraries or rapidly type in code and before executing even once, you are looking at 50 lines of code. When finally the code is run, you are totally lost and still stuck trying to understand line 3. This causes many students to give up learning Neural Networks.This course is different! It starts with the absolute beginning and each topic is a continuation of a previous example. This way, you will learn neural networks from the ground up, step by step.What can you do after this course?You understand neural network concepts and ideas, like back propagation and gradient descent.You are able to build a neural network in any programming language of choice, without the help of frameworks and libraries.You understand how to better configure the network by plugging in different cost functions and adding hidden layers.TopicsLinear regressionCost functionsBiasMultiple inputsNormalisationGradient descentClassificationActivationMulti-class classificationNon-linear dataHidden layersDuration3 hour video time. This course has no exercises.The teacherThis course is taught by Loek van den Ouweland, a senior software engineer with 25 years of professional experience. Loek is the creator of Wunderlist for windows, Microsoft To-do and Mahjong for Windows and loves to teach software engineering.Students of this course tell me:* * * * * "Great, simple explanations. Perfect for beginners that have little pre knowledge of the topic."* * * * * "Straight to the point starting with the foundations."* * * * * "Clearly explained step by step how Neural Networks work and can be developed in a pure development language of choice without the usage of any external package.."
