The SDML book club started discussing Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow, 2nd Edition by Aurelien Geron in October 2021.
Below are the notes/slides and meetup recordings. All of the videos are on Ryan's YouTube channel: https://www.youtube.com/c/ITConnected/videos
(Please scroll down for Part I of the book)
Chapter 10: Introduction to Artificial Neural Networks with Keras
Meetup: Chapter 10 event
Discussion leader: Jerry Kurata
This meetup will kick off our hands-on series about artificial neural networks and deep learning,
starting with an introduction to neural networks, and how to build them using Keras and TensorFlow 2.
Slides, the Notebook on Jerry's GitHub,
and video recording
Chapter 11: Training Deep Neural Networks
Meetup: Chapter 11 meetup
Discussion leader: Robert Kraig
This week we will drill into some of the issues in training deep neural networks, including taking advantage of transfer learning, and using different optimizers, learning rates, and regularization tools.
Slides and video
Chapter 12: Custom Models and Training with TensorFlow
Meetup: Chapter 12 session
Discussion leader: George Zoto
This session will provide an overview of TensorFlow and look into how you can do customizations when Keras doesn't have exactly what you need predefined for you.
Slides and the links to resources, including George's Jupyter notebook, are on slide 6. Video recording is on YouTube.
Chapter 13: Loading and Preprocessing Data with TensorFlow
Meetup: Chapter 13 event
Discussion leader: Kirk Parker
Real world deep learning requires reading huge datasets efficiently. Preprocessing the data - including encoding and normalizing - is often necessary as well.
This session will discuss the capabilities built into Keras and TensorFlow to handle these needs.
Slides and video
Chapter 14: Deep Computer Vision Using Convolutional Neural Networks
Meetup: Chapter 14 session
Discussion leader: George Zoto
Convolutional Neural Networks (CNNs) brought forth the revolution of deep learning.
This week we will explain convolutional layers and how they are used in the CNN architecture for several different computer vision tasks.
Slides (note that George's terrific list of resources is on slides 18 & 19)
and video
Chapter 15: Processing Sequences Using RNNs and CNNs
Meetup: Chapter 15 meetup
Discussion leader: Steven Fouskarinis
This session examines RNNs and other building blocks for dealing with sequential data.
Next meeting we will go into Natural Language Processing in further detail.
Slides and video
Chapter 16: Natural Language Processing with RNNs and Attention
Meetup: Chapter 16 event
Discussion leader: Vibhu Sapra
This session dives deeper into natural language processing tasks and introduces the revolutionary transformer architecture.
Slides
Chapter 17: Representation Learning and Generative Learning Using Autoencoders and GANs
Meetup: Chapter 17 session
Discussion leader: Robert Kraig
This week we will discuss encoding images into and generating images from lower dimensional spaces.
Slides and video recording
Chapter 18: Reinforcement Learning
Meetup: Chapter 18 event
Discussion leader: Rajinder Mavi
This meeting we will do a whirlwhind tour of reinforcement learning.
If you are interested in going deeper, check out the book club series we did on reinforcement learning.
video
Chapter 19: Training and Deploying TensorFlow Models at Scale
Meetup: Chapter 19 meetup
Discussion leader: Nidhin Pattaniyil
This final session we will discuss training models on GPUs or multiple GPUs, as well as how to deploy and serve models.
Slides and video
Chapter 1: The Machine Learning Landscape
Here is the full Meetup event description.
We met other people in small breakout rooms and discussed chapter 1 of the HOML book.
Notes and the Video of chapter 1.
Chapter 2: End-to-End Machine Learning Project
Meetup: Chapter 2 event
Discussion leader: Ian Cook
Notes and the Video of first session. We will finish this chapter next week.
Chapter 3: Classification
Meetup: Week 3 event
Discussion leader: Vibhu Sapra
We will finish talking about the end-to-end ML process, then review classification problems.
Slides and Video of the combo session.
Chapter 4: Training Models
Meetup: Chapter 4 event
Discussion leader: Anne Knowles
We will discuss several foundational techniques in ML including linear regression, logistic regression, and gradient descent.
We will also look at how models are trained, the bias/variance trade-off, and regularization.
Slides and Chapter 4 video
Chapter 5: Support Vector Machines
Meetup: Chapter 5 event
Discussion leader: Bhanu Yerra
This week we will discuss the core concepts of Support Vector Machines, how to use them, and how they work.
Slides and video
Chapter 6: Decision Trees
Meetup: Chapter 6 event
Discussion leader: Glenn Miller
This week's topic will be decision trees.
Slides and video
Chapter 7: Ensemble Learning and Random Forests
Meetup: Chapter 7 event
Discussion leader: Nidhin Pattaniyil
This week we will discuss the general case of ensemble learning, where we use multiple models,
and the specific algorithms of Random Forests and Gradient Boosted Decision Trees.
Slides and video
Chapter 8: Dimensionality Reduction
Meetup: Chapter 8 event
Discussion leader: Harold Anderson
Dimensionality reduction techniques are used to cut down the number of features when training models,
and are also very useful for visualizing high-dimensional data.
Slides and video
Chapter 9, Part 1: Clustering Algorithms
Meetup: Chapter 9, Part 1 meetup
Discussion leader: Kalika Curry
This meeting will begin our discussion of unsupervised learning techniques. We start with clustering algorithms.
Video for part 1
Chapter 9, Part 2: Gaussian Mixture Models
Meetup: Chapter 9, Part 2 meetup
Discussion leader: Kalika Curry
This meeting we wrap up our review of unsupervised learning techniques with Gaussian mixture models.
Slides for both parts of chapter 9 and video for part 2
Jupyter notebooks with code from Hands-On Machine Learning is available in this repo: https://github.com/ageron/handson-ml2.