External Learning Resources

This document serves as a compilation of external learning resources on the topics of AI & ML (Primarily technical learning resources, with emphasis on free/open source tools and content)

Conceptual Introduction to AI (no prerequisites)

Resources for increasing fluency and understanding of AI applications and impact, no programming or math needed.

Books/Articles

• You look like a thing and I love you : How Artificial Intelligence Works and Why It’s Making the World a Weirder Place (Accessible Introduction to AI) - Also available digitally and physically at PCC and the Multnomah County library.

Courses

• AI for Everyone - 6hr overview course with no technical prerequisites
  • Lectures are broken up into a lot of short clips,could be useful for illustrative purposes.

Videos/Channels

Technical Introduction to AI

(ranging from beginner to advanced; prerequisites noted)

There are many free and accessible AI courses, tools and libraries that allow you to leverage AI models & technology with or without getting into the nitty gritty of the math and model development. Many of these resources assume minimal programming knowledge and high-school level maths. Unlike the previous resources, these delve into technical topics and may include problem-solving activities. Some of these learning resources have more advanced prerequisites (which will be noted - see AI prerequisites section for where to gain these skills).

Books/Articles

Intro readings

• Make your own Neural Network (Friendly, step-by-step technical introduction to neural networks & python)
  • Book exercises/tutorial github
  • Accompanying Blog
• http://neuralnetworksanddeeplearning.com/ (free digital only text - intros NNs & Python)
  • Code Samples for text

Intermediate readings

• The Hundred-Page Machine Learning Book - Compact overview of Supervised/Unsupervised Learning with code tutorials (Presumes math/programming knowledge - good compact reference)
• Deep Learning (Comprehensive Technical Textbook style text with math focus)
  • Deep Learning Textbook Resource Page (Includes Digital Version of Text , tutorials, & lecture videos, discussion group, etc)

Advanced (Math/Prog. knowledge expected)

Courses

Intro courses

• Andrew Ng’s Machine Learning Specialization (3 Course Series, ~33hrs each)
  • Assumes highschool math (Algebra concepts, etc.) No higher level math and only basic coding experience required (See Ada’s 1-week Python Course)
• MIT’s Intro Deep Learning
  • Introduction to deep learning methods with applications to computer vision, natural language processing, biology, and more
  • Jumps straight into applications, best to orient to theory elsewhere, but is more technically accessible than courses below
  • Prerequisites: very elementary knowledge of linear algebra and calculus. How to multiply matrices, take derivatives and apply the chain rule. Familiarity in Python is a big plus as well. The course will be fairly beginner friendly, even if you don’t know all the prereqs, you can learn as you go.
  • Time Commitment: (Originally presented as a 2-week bootcamp, ~30-40hrs)
• Harvard’s CS50 Introduction to Artificial Intelligence with Python
  • Introduces both concepts and foundational AI algorithms through project-based learning.
  • Prerequisites: CS50P (good Python intro course) or prior python programming experience
  • Time Commitment: 70 -200 hours
• Google’s Machine Learning Crash Course
  • Introduces practical application of ML to problem solving with Tensorflow API
  • Prerequisites: Intro to Machine Learning Problem Framing (1hr), Basic Python, NumPy & Pandas knowledge (python libraries)
  • Time Commitment: 15hrs
• MIT’s Deep Learning for Art, Aesthetics, & Creativity

Additional Non-Traditional Courses

• Machine Learning Bootcamp - 4 month cohort model or self-paced option, all resources free. Assumes no math exerience, but some programming experience (~1yr recommended). Based off of this book: Machine Learning Bootcamp
• 60 Days of Data Science and Machine Learning with Project Series - Medium post series, covers basics of python as well (prereq for following)
• 60 Days of Deep Learning with Projects Series - Medium Post Series (Uses Jupyter Notebooks + Google Colab) - Must Pay for Medium membership

Intermediate Courses

• UPenn’s Intro to Artificial Intelligence for CIS Majors
  • Prerequisites: CIS 121, Data Structures & Algorithms, Programming Experience
  • Required Text: Artificial Intelligence: A Modern Approach by Russel and Norvig
  • Standard 1 Semester 4 Credit Course

Advanced Courses

• Intro to Deep Learning NYU - Graduate level Deep Learning course. Focus on Energy-Based-Models.
  • Prequisites: presuming prior statistical or ML knowledge, linear algebra, etc.
  • Time Committment: 5hrs/wk, 14wks of material (60hrs)
  • Under CCA License (we can share/remix with attribution)

Videos / Channels

• Harvard Medical AI Lecture series

Blogs/Websites of Interest

Additional Teaching/Learning Resources (varying prereqs)

• Machine Learning for Artists (ITP tutorials, example projects, etc. using ML5.js and Python)
• DeepMind educational Github (Co-lab notebook based tutorials on a range of topics)
  • Also includes CCA licensed teaching resources for Introduction to Machine Learning
  • Some Python experience helpful
• Kaggle: Data Science oriented machine learning tutorials and challenges (introduces python, data science libraries, ml libraries, geospatial analysis etc. and provides a wide range of audio, textual, visual data sets for testing and validation).
  • Highly interactive (uses Jupyter notebooks for IDE learning)
  • Each tutorial takes approximately 4 hrs to complete (some stack)
    • Gets you started, but not that detailed - recommend additional learning elsewhere
  • good source of data sets
• https://github.com/josephmisiti/awesome-machine-learning (Meta list of ML learning resources, models, libraries, etc.)
• https://github.com/openhackathons-org/gpubootcamp (Series of Jupyter Notebook tutorials designed to prepare students for participation in AI/GPU hackathons - topics included intro to AI model training, physics informed AI, etc. )
  • Best for folks with a math/science background who want to learn to utilize ML & GPU acceleration.

Software & Hardware Tutorials & Trainings

Tutorials and trainings that address a specific application or set of applications - may require the use of specific hardware or paid software.

• Tiny Machine Learning Open Education Initiative (Repo of Courses and Instructional Support Materials)
  • Requires https://store-usa.arduino.cc/products/arduino-tiny-machine-learning-kit
• Edge Impulse University Program - Online IDE & educational resources (including free hardware if requested) for building and deploying microcontroller based machine learning.
  • Courseware Gitub
• NVIDIA Deep Learning Institute - Range of Online Courses around various ML tools & hardware (Free and Paid content, Certifications Available)
  • Self-guided learning and live trainings available (certificate courses average 8hrs each)
  • Most trainings expect some familiarity with a base programming language (Python, C++, etc.) Look for “Beginner” courses for the lowest technical barrier.
  • Range from 10 minute tutorials to 8 hour courses. Learn through implementation.

Additional Courses and References

• 100 Lectures on Machine Learning from Mark Schmidt at UBC
• Google's Machine Learning Engineer Learning Path - collection of courses, skill badges, etc. required for certification
• Machine Learning Interview Guide - perspective/study plan for Machine Learning focused technical interviews

Things found during the course of learning about AI, that may not be about AI

Tools used in conjunction with AI, interesting theories or related domains of study/etc

• Creative coding for interactive art (ITP) - range of notes and resources, some of which touch on machine learning