|
所在平台: Udemy |
课程主页: https://www.udemy.com/course/building-self-driving-cars-in-python-from-scratch/
课程评论：没有评论
课程名称：从零开始使用遗传算法构建自动驾驶汽车 课程概述：自动驾驶汽车的实验可以追溯到1939年，但真正的自主驾驶汽车始于1980年代大学的研发。当时，一辆无驾驶员的梅赛德斯-奔驰在1984年以每小时130公里的速度行驶！这一项目获得了高达7.49亿欧元的资金支持。如今，构建人工智能不再需要如此庞大的预算，只要有一台安装了Python的计算机就可以了！但要如何开始构建自动驾驶汽车的AI呢？本课程将带领你从零开始学习构建神经网络和遗传算法，而不是依赖隐藏许多细节的框架。你将创建一个训练自动驾驶汽车的程序，学习并组合所有必需的构建模块，不久后汽车便能自主行驶。学习AI的方式只有一种，就是选择一个项目并开始构建，这正是你在本课程中所要做的！ 目标受众：本课程特别适合以下开发者： - 想利用基础Python技能编程自动驾驶汽车的开发者。 - 希望通过从头开始构建来理解神经网络和遗传算法的开发者。 挑战：人工智能对许多开发者来说是一个黑箱，许多AI框架隐藏了必要的细节，使得理解各个组件的工作原理变得复杂。本课程的解决方案是从零开始构建，学习如何创建和组合遗传操作员，以及如何调整属性以优化结果。本课程将从一个空白脚本开始，逐步展示创建学习如何在赛道上驾驶的自主汽车所需的每一个步骤。一旦掌握了遗传算法的构建模块，你就可以在未来的项目中加以应用！ 课程完成后你可以： - 定义遗传算法能解决的问题。 - 从零开始构建神经网络和遗传算法。 - 采用遗传算法解决可以用遗传算法解决的任何问题，并复用在本课程中创建的代码。 课程主题： - AI简介：神经网络和遗传算法 - 汽车机械：创建窗口、绘制背景和汽车、控制汽车，理解赛道信息 - 神经网络：输入、输出、传感器、激活、前馈 - 遗传算法：适应度、染色体、选择、交叉和变异 - 挑战：滑行的汽车、存储汽车大脑、保持在道路中间及测试驾驶 课程时长：视频时间2小时，包含跟打的总时长约为6小时。 讲师简介：本课程由Loek van den Ouweland教授，他是一位拥有25年专业经验的高级软件工程师。他是Wunderlist for Windows、Microsoft To-do和Windows版麻将的创建者，非常热爱软件工程的教学。
Self-driving car experiments go back to 1939. But it took until the 1980's when universities started to create true, autonomous cars. In Munich, a driverless Mercedes-Benz was going a whopping 130KM/H in 1984!That is 81 miles per hour. And without crashing! The project received an astronomical funding of 749,000,000 Euros.These days, you don't need such budgets for artificial intelligence. All you need is a computer with Python on it! But where to start to build the AI for self-driving cars?In this course you learn to build Neural Networks and Genetic Algorithms from the ground up. Without frameworks that hide all the interesting stuff in a black box, you are going to build a program that trains self-driving cars.You will learn and assemble all the required building blocks and will be amazed that in no time cars are learning to drive autonomously. There is only one way to learn AI and that is to just pick a project and start building. That is what you are going to do in this course!Target audienceDevelopers who especially benefit from this course, are:developers who want to use their basic Python skills to program self-driving cars.developers who want to understand Neural Networks and Genetic Algorithms by building them from the ground up.ChallengesArtificial Intelligence is a black box to many developers. The problem is that many AI frameworks hide the details you need to understand how all the individual components work. The solution is to build things from the ground up and learn to create and combine genetic operators and what properties you can change to optimize the result. This course starts with an empty script and shows you every step that is needed to create autonomous cars that learn how to drive on tracks. Once you have seen the building blocks of a Genetic Algorithm, you can use them in your future projects!What can you do after this course?define what problems can be solved with Genetic Algorithmsbuild Neural Networks and Genetic Algorithms from the ground uptake any problem that can be solved with genetic algorithms and solve it by re-using the code you created in this courseTopicsAI Introduction: Neural Networks and the Genetic AlgorithmCar mechanics: Creating a window, drawing backgrounds and cars, controlling the car. Understanding track informationNeural Network: Inputs, outputs, sensors, activation, feed forwardGenetic Algorithm: Fitness, Chromosomes, Selection, Cross over and MutationChallenges: Slipping cars, Store the car brain, Stay in the middle of the road and Test DrivesDuration2 hours video time, 6 hours including typing along.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.
