CS230 Deep Learning is an introductory course on deep learning that will offer a broad overview of the field. This course is a part of the Stanford AI Lab’s Autumn 2018 course schedule.
Check out our new video:
Artificial intelligence (AI) is one of the hottest areas in all of tech right now. Westworld and Blade Runner 2049 piqued our curiosity about what life with thinking machines might be like, while new Research from Google DeepMind seem to show that AI might be able to outsmart us at some of our most complex tasks, like creating works of art or even developing new scientific theories.
This course will cover Cutting Edge research in Deep Learning. We will start with a recap of linear models and backpropagation, and discuss recent developments in training Deep Neural Networks. The rest of the course will be focused on modern techniques for training Deep Neural Networks. This will include topics such as Optimization Methods (SGD, momentum, RMSProp, Adam), Regularization (dropout, batchnorm), Convolutional Neural Networks, Recurrent Neural Networks, Generative Models (Autoencoders, VAEs, GANs), Reinforcement Learning and lots more.
This course will cover the fundamental Deep Learning methods that are currently used by industry, including convolutional networks, recurrent networks, generative models, and reinforcement learning. The course will focus on the practical aspects of Deep Learning and will include hands-on assignments and projects. At the end of the course, students will be able to build Deep Learning systems that can achieve state-of-the-art performance on a variety of tasks.
This autumn, CS230 will be taught by Professors Percy Liang and Andrew Ng.
This course is an advanced course on deep learning, covering both the theory and applications of deep neural networks. It is open to Stanford undergraduates, graduate students, postdoctoral scholars, and visiting scholars.
Prerequisites: CS229 or equivalent. Knowledge of basic machine learning (e.g., CS221 or CS229) and deep learning (e.g., CS230) is required.
There are no exams in this course. However, there will be several quizzes and homework assignments to test your understanding of the material. The final project will be a group project, with each group required to design and train a deep neural network to solve a real-world problem of their choice.
The course website will be updated with lecture videos, slides, and handouts after each lecture. Please check the website regularly for announcements and updates.
This course will have three assignments, released approximately every two weeks. Assignments will consist of both written and programming components, and will be based on material from lectures and readings.
$\therefore$ **Course projects:** You will work on 3 projects in groups of 4-5 students during the quarter. The first project will be an implementation of a published research paper, to get you up to speed with the relevant deep learning libraries and frameworks. For the second project, you will have the opportunity to design and train your own state-of-the-art models on an original dataset of your choice (potentially including your own data collection). For the third project, you can either (i) do an even more open-ended research/development project of your own design, or (ii) take on an additional comprehensive paper implementation (in this case working in pairs).
This quarter we will be discussing a wide range of topics in deep learning. Below is a tentative schedule for the course. Please note that this schedule is subject to change.
Week 1: Introduction and course overview
Week 2: Supervised learning (Regression and Classification)
Week 3: Neural networks Part I
Week 4: Neural networks Part II
Week 5: Convolutional neural networks
Weeks 6-7: Recurrent neural networks
Week 8: Unsupervised learning
Week 9: Generative models
Weeks 10-11: Reinforcement learning
Week 12: Special Topic (adversarial learning, domain adaptation, structured prediction, etc.)
For those of you who started the CS230 Deep Learning course this autumn, we hope that you are all finding the course engaging and beneficial so far. We would greatly appreciate it if you could take a few moments to provide feedback on the course by filling out this form: https://goo.gl/forms/sOIe0Tbvjnm0Z8Gb2. Your feedback is very important to us and will help us improve the course for future students.
The CS230 Staff
This article concludes our introduction to the Stanford CS230 deep learning course. We have reviewed the course structure, the topics covered, and the skills that you will need to succeed. In addition, we have provided some advice on how to make the most of your learning experience.
We hope that this article has given you a better understanding of what to expect from the CS230 course. If you have any questions, please don’t hesitate to ask in the comments section below.
Keyword: Stanford CS230 Deep Learning: What to Expect in Autumn 2018