Introduction to Artificial Intelligence and Deep Learning for Engineers and Scientists

工程科學之人工智慧與深度學習導論

Course description

Artificial Intelligence (AI) already impacts the life of essentially all internet users, but at the same time, it is Science Fiction. Deep Learning (DL) has since the ImageNet LSVRC-2012 contest established itself as one of the core technologies driving the third industrial revolution with many commercial applications. The success of the current wave of Artificial Narrow Intelligence (ANI) is due to:
  • big labelled data,
  • GPU accelerated distributed computing,
  • open source software, and
  • algorithms.

In this course, we will discuss what AI and DL is, their applications and implications from an automation perspective, the importance of ethics and danger of algorithm bias, and the methods that make AI possible, with a focus on Deep Neural Networks. You will also get to apply a Neural Network to identify numbers in images, using Python and TensorFlow. No previous knowledge of AI, Python, or TensorFlow is needed.

人工智慧(AI)已影響基本上所有的網路使用者，然而同時它也還是科幻小說的範疇。自從ImageNet LSVRC-2012競賽以來，深度學習(DL)已然成為一驅使第三次工業革命與其不可勝數的商業應用的核心科技，能有現在人工智慧(ANI)的成功，主要歸功於：

  • 大標記數據
  • GPU加速之分散式運算
  • 開源軟體
  • 演算法

在本課程中，我們將會討論人工智慧、深度學習、應用、自動化之觀點、倫理道德、演算法偏差以及實現人工智慧的方法。此外，本課程會有利用Python與Tensorflow神經網路判讀數字圖片的練習，課程前並不需要Python與Tensorflow的先備知識。

Course Objectives

  • Define Artificial Intelligence
  • Know some applications where AIs outperform humans
  • Understand some major implications of AI
  • Ability to critically examine AI news and claims
  • Understand the importance of ethics and algorithm bias
  • Explain some methods that make AI possible
  • Explain what an Artificial Neural Network is
  • Explain how Convolution works
  • Explain how Backpropagation works
  • Apply a Neural Network to recognise numbers

Instructor

 Ass. Prof. Torbjörn E. M. Nordling, Ph.D.

Teaching Strategies

How this course is going to be taught

  • Using online teaching 60% 60%
  • Case study 20% 20%
  • Presentation 10% 10%
  • Video/Music Appreciation 10% 10%

Course material

Short videos explaining a topic with companion lecture notes, suggested reading, exercises, and/or example code. All material is or will be freely available online. All material produced specifically for this course will be under the Creative Commons Attribution 4.0 International (CC BY 4.0) license with attribution to Prof. Nordling, Nordling Lab.

Course outline (2020)
  1. Success stories of Artificial Intelligence
  2. What is AI, Deep Learning, and Machine Learning?
  3. Introduction to Python
  4. Introduction to Python
  5. Introduction to Neural Networks and Modeling
  6. Introduction to TensorFlow
  7. Introduction to data analysis and Machine Learning
  8. Gradient descent, Backpropagation, and loss functions
  9. Introduction to Model validation
  10. Application of Deep Learning
  11. Ethics and the danger of algorithm bias
  12. Application of Deep Learning
  13. Automation and the future of work
  14. Application of Deep Learning
  15. Current AI research
  16. Application of Deep Learning
  17. Application of Deep Learning
  18. Final exam and group presentations
References
  1. Goodfellow, I., Bengio, Y. & Courville, A., 2016. Deep learning, Cambridge, MA, U.S.A.: MIT Press. Available at: http://www.deeplearningbook.org.
  2. O’Neil, C., 2016. Weapons of Math Destruction: How Big Data Increases Inequality and Threatens Democracy, Crown Random House. Available at: https://weaponsofmathdestructionbook.com/.
  3.  Jordan, M.I., 2018. Artificial Intelligence — The Revolution Hasn’t Happened Yet. Medium. Available at: https://medium.com/@mijordan3/artificial-intelligence-the-revolution-hasnt-happened-yet-5e1d5812e1e7.
  4. Purdy, M. & Daugherty, P., 2016. Why artificial intelligence is the future of growth, Available at: https://www.accenture.com/us-en/insight-artificial-intelligence-future-growth.

Course policy

  1. Participation 30%
    Counted based on a weekly diary entry committed to the GIT repository before 24:00 every Sunday following these instructions. Each student need to write down what he/she has learnt during the week. This is counted from the third week onwards.
    In the master and doctoral version of this course (N182200 / ME7122), each student need to read and summaries (in 3-5 bullet points) one recent open access scientific article on training of Deep Neural Networks using supervised learning. The summary must be included as the last entry in the diary file with header # Article summary #.
  2. Group project 40%
    Graded based on final presentation during a synchronized (or physical) lecture, which should contain a demonstration of how well the trained Deep Neural Network performs on an independent test data set. A Jupyter notebook with both the training and prediction code, as well as the model specification and final model (saved in HDF5 format), need to be committed to the GIT repository in accordance with these instructions.
  3. Final exam 30%
    The exam will be conducted online using Google forms. The final exam will consist of questions with multiple choice.

Course website

Licensed under CC BY 4.0, except logos and material from published articles.