Learn AI principles and code through a 12-week, 24-lesson path
This is a journey for those new to artificial intelligence, following 24 lessons over 12 weeks to understand concepts and run actual code. The course covers a wide range of topics, from the basics of neural networks and deep learning to computer vision and natural language processing, with each topic structured around executable notebook files and exercises. Readers do not just read theory; they gain practical understanding by running code to see how AI works.
A learning flow from basics to modern models
The course begins with the history of AI and symbolic approaches before moving to neural networks and deep learning, the core of modern AI. It teaches neural network structures for handling images and text using two major frameworks, TensorFlow and PyTorch. It then covers less common approaches like genetic algorithms and multi-agent systems, as well as AI ethics, providing a broad perspective. While explanations of the latest models may be somewhat limited, the course maintains sufficient depth to grasp core concepts.
Practice-oriented learning materials
Each lesson consists of pre-reading materials and executable Jupyter Notebooks. The notebook files include theoretical explanations, allowing you to proceed by choosing either PyTorch or TensorFlow. Some topics offer labs that solve specific problems, enabling you to apply what you have learned immediately. Additionally, links connect to Microsoft Learn modules, opening a path to explore related topics in greater depth.
Areas this course does not cover
This curriculum does not cover business cases, classical machine learning, practical AI applications based on Azure, specific cloud frameworks, conversational AI, or the deep mathematics behind deep learning. These topics are better suited to other learning paths on Microsoft Learn or books by authors such as Ian Goodfellow. Therefore, this course focuses on the basic principles of AI and coding practice, while business application or operating specific cloud services requires separate study.
Preparation before starting
You must prepare the necessary tools through the first lesson, which sets up the development environment. When cloning the repository, translation files significantly increase the download size, so if you do not need translations, it is recommended to use sparse checkout to exclude unnecessary files. If you are completely new to AI, it is recommended to start with beginner-friendly examples like 'Hello AI World' before entering the official curriculum. You can also join the official Discord server to communicate with other learners and receive help.
When to choose this course
This course is suitable for beginners who want to grasp basic AI concepts and coding practice simultaneously. It is particularly useful for learners who want to build neural networks directly by choosing either TensorFlow or PyTorch. However, if your goal is to follow the latest research trends or operate specific cloud services in practice, this course alone may be insufficient. If you want a deep mathematical background, you should read additional reference materials. If your learning goals are aligned with understanding basic principles and running code, this 12-week course is a good starting point.
