Introduction to AI & ML
Class & Curriculum Outline
Learn machine learning, with us! A joint collaboration with ACM AI Outreach.
We present a year-long, modern machine learning class led by UCLA students! Students will discover and explore the computational and mathematical tools behind artificial intelligence and machine learning! In a single year, they will advance from understanding simple AI models that predict weather to analyzing the complex AI systems that power self-driving cars.
Throughout the year, students will have numerous opportunities to learn and code with Python. Python is one of the most popular programming languages in use today, with wide-ranging applications in data processing, web development, and machine learning. Last year, our students used Python to train AI models that predict stock market prices, guess the popularity of a song on Spotify, and detect handwritten digits.
Learning machine learning does not occur in a vacuum! We encourage our students to think critically about the applications and ethics of AI. In the past, we’ve brainstormed which types of AI are best-suited to tackle facial recognition and estimate house prices. Furthermore, we’ve held in-depth discussions on the ethics of AI-powered self-driving cars and what it actually means for AI to be racist.
#1 What is AI?
You’ve probably heard of Artificial Intelligence in the news. But what exactly is AI?
Deep learning has revolutionized AI! But, what exactly is it, and how is it different from AI and machine learning?
As mad as LeBron at JR Smith for costing the Cavaliers Game 1 Of the 2018 NBA Finals? Channel that anger into learning about supervised learning, training, and testing!
Lines are powerful! (You may take a look at the Colab now, but it will make more sense by Lesson 4d.)
KK Slider teaching you how to multiply matrices?! Heck yeah!
#4c Gradient Descent
You heard right! Machine learning models train by taking hikes down steep mountains. After all, the decision trees are quite beautiful.
Lines are even more powerful when you know TensorFlow!
Let’s train a model that can classify an image as being of either a dog or cat!
Hey, baes, let’s leverage Bayes’ Theorem to build a spam filter!
Using probability as a shovel, we’ll dig a little deeper into binary cross-entropy loss (you know, the thing that we optimize to train logistic regression models).
Ooh, let’s demystify neural networks! (And, no, they do not work like the brain does.)
Gradient descent is throwing a party, and we’re all invited! Get ready to meet its family and friends.
Images, filters, and networks, oh my! We’ll break down convolutional neural networks without any convolutions. (slides are in progress)
These are our favorite AI/ML resources - they're all awesome!
ACM AI Resources
- You Belong in AI! Podcast
- ACM AI Blog
- Plotting Data Notebook
- MNIST Dataset with Keras Notebook (Archer School Event)
Intro to Deep Learning
- Coursera: Deep Learning Specialization
- The Deep Learning Textbook
- 3Blue1Brown: Neural Networks
- Udacity: Intro to Deep Learning with TensorFlow
Tensorflow and Pytorch
- UC Berkeley: CS294 Tensorflow Tutorial
- UC Berkeley: CS285 (Deep Reinforcement Learning) Resources
- Pytorch: Getting Started with Pytorch