← Back
Hands_on_ML.pdf
Q&A History
Q: Give me summary of page 109
I don't have enough information to answer that. The context provided does not contain content from page 109 of the book. The excerpts shown are from the introduction (Chapter 1), Appendix A (Thank You section), and contact/publishing information pages, but not from page 109 specifically.
3 source chunks
scientist should know by heart. It will be a high-level overview (the only chapter
without much code), all rather simple, but you should make sure everything is
crystal-clear to you before continuing to the rest of the book. So grab a coffee and let’s
get started!
3
Download from finelybook www.finelybook.com
If you already know all the Machine Learning basics, you may want
to skip directly to Chapter 2 . If you are not sure, tr
y to answer all
Appendix A.
Thank You!
Before we close the last chapter of this book, I would like to thank you for reading it
up to the last paragraph. I truly hope that you had as much pleasure reading this book
as I had writing it, and that it will be useful for your projects, big or small.
If you find errors, please send feedback. More generally, I would love to know what
you think, so please don’t hesitate to contact me via O’Reilly, or through the ageron/
handson-ml GitHub project.
707-829-0104 (fax)
We have a web page for this book, where we list errata, examples, and any additional
information. Y ou can access this page at http://bit.ly/hands-on-machine-learning-
with-scikit-learn-and-tensorflow .
To commen
t or ask technical questions about this book, send email to bookques‐
[email protected].
For more information about our books, courses, conferences, and news, see our web‐
site at http://www.oreilly.com.
Find us on Facebook: http://facebook.com/oreilly
March 31, 2026, 6:44 p.m.
Q: What is this book about
Based on the context provided, this book is about **Machine Learning**.
The context mentions that it includes chapters covering Machine Learning basics and concepts that "a scientist should know by heart," and it references chapters with code examples related to Machine Learning projects. The book appears to be a hands-on guide to Machine Learning, as suggested by the GitHub project reference "ageron/handson-ml."
However, I don't have enough detailed information in the provided context to give you a more specific description of all the topics covered or the full scope of the book.
3 source chunks
scientist should know by heart. It will be a high-level overview (the only chapter
without much code), all rather simple, but you should make sure everything is
crystal-clear to you before continuing to the rest of the book. So grab a coffee and let’s
get started!
3
Download from finelybook www.finelybook.com
If you already know all the Machine Learning basics, you may want
to skip directly to Chapter 2 . If you are not sure, tr
y to answer all
Press, FT Press, A press, Manning, N
ew Riders, McGraw-Hill, Jones & Bartlett, and
Course Technology, among others.
For more information, please visit http://oreilly.com/safari.
How to Contact Us
Please address comments and questions concerning this book to the publisher:
O’Reilly Media, Inc.
1005 Gravenstein Highway North
Sebastopol, CA 95472
800-998-9938 (in the United States or Canada)
707-829-0515 (international or local)
707-829-0104 (fax)
Appendix A.
Thank You!
Before we close the last chapter of this book, I would like to thank you for reading it
up to the last paragraph. I truly hope that you had as much pleasure reading this book
as I had writing it, and that it will be useful for your projects, big or small.
If you find errors, please send feedback. More generally, I would love to know what
you think, so please don’t hesitate to contact me via O’Reilly, or through the ageron/
handson-ml GitHub project.
March 31, 2026, 6:43 p.m.
Q: github
Based on the context provided, here's what I can tell you about GitHub:
1. TensorFlow is one of the most popular open source projects on GitHub
2. You can file bugs and feature requests for TensorFlow through GitHub
3. There is an "awesome-tensorflow" resource list available at https://github.com/jtoy/awesome-tensorflow that showcases great projects being built on top of TensorFlow
The context mentions GitHub as a platform where TensorFlow is hosted and where the community can report issues and contribute to the project.
3 source chunks
ulations if you wanted.
developed by the Google Brain team and it powers man
y of Google’s large-scale serv‐
ices, such as Google Cloud Speech, Google Photos, and Google Search.
Figure 9-2. Parallel computation on multiple CPUs/GPUs/servers
When TensorFlow was open-sourced in N ovember 2015, there were already man
y
popular open source libraries for Deep Learning (Table 9-1 lists a few), and to be fair
most of TensorFlow’s features already existed in one library or another. Nevertheless,
Find us on Facebook: http://facebook.com/oreilly
Follow us on Twitter: http://twitter.com/oreillymedia
Watch us on Y ouTube: http://www.youtube.com/oreillymedia
Acknowledgments
I would like to thank my Google colleagues, in particular the Y ouTube video classifi‐
cation team, for teaching me so much about Machine Learning. I could never have
started this project without them. Special thanks to my personal ML gurus: Clément
proving it. It is one of the most
popular open source projects on GitHub, and more and more great projects are
being built on top of it (for examples, check out the resources page on https://
www.tensorflow.org/, or https://github.com/jtoy/awesome-tensorflow). To ask
technical questions, you should use http://stackoverflow.com/ and tag your ques‐
tion with "tensorflow". Y ou can file bugs and feature requests through GitH
ub.
For general discussions, join the Google group.
March 10, 2026, 11:09 a.m.
Q: Does author mention their friend?
Based on the context provided, the author mentions their **father-in-law, Michel Tessier**, who is described as a "former mathematics teacher and now a great translator of Anton Chekhov." The author thanks him for helping with mathematics, notations, and reviewing the linear algebra Jupyter notebook.
However, the context does not clearly indicate whether Michel Tessier is considered a "friend" specifically. The author refers to him as "my father-in-law," which is a family relation rather than explicitly calling him a friend.
3 source chunks
Thanks as well to Eddy Hung, Salim Sémaoune, Karim Matrah, Ingrid von Glehn,
Iain Smears, and Vincent Guilbeau for reviewing Part I and making many useful sug‐
gestions. And I also wish to thank my father-in-law, Michel Tessier, former mathe‐
matics teacher and now a great translator of Anton Chekhov, for helping me iron out
some of the mathematics and notations in this book and reviewing the linear algebra
Jupyter notebook.
Press, FT Press, A press, Manning, N
ew Riders, McGraw-Hill, Jones & Bartlett, and
Course Technology, among others.
For more information, please visit http://oreilly.com/safari.
How to Contact Us
Please address comments and questions concerning this book to the publisher:
O’Reilly Media, Inc.
1005 Gravenstein Highway North
Sebastopol, CA 95472
800-998-9938 (in the United States or Canada)
707-829-0515 (international or local)
707-829-0104 (fax)
Courbet, Julien Dubois, Mathias Kende, Daniel Kitachewsky, James Pack, Alexander
Pak, Anosh Raj, Vitor Sessak, Wiktor Tomczak, Ingrid von Glehn, Rich Washington,
and everyone at Y ouTube Paris.
I am incredibly grateful to all the amazing people who took time out of their busy
lives to review my book in so much detail. Thanks to Pete Warden for answering all
my TensorFlow questions, reviewing Part II, providing many interesting insights, and
March 10, 2026, 11:09 a.m.
Q: Tell me about author of this book?
I don't have enough information to answer that. The context provided is from Appendix A of a book, which contains acknowledgments and contact information, but it does not include the author's name or biographical information about them.
3 source chunks
Thanks as well to Eddy Hung, Salim Sémaoune, Karim Matrah, Ingrid von Glehn,
Iain Smears, and Vincent Guilbeau for reviewing Part I and making many useful sug‐
gestions. And I also wish to thank my father-in-law, Michel Tessier, former mathe‐
matics teacher and now a great translator of Anton Chekhov, for helping me iron out
some of the mathematics and notations in this book and reviewing the linear algebra
Jupyter notebook.
Press, FT Press, A press, Manning, N
ew Riders, McGraw-Hill, Jones & Bartlett, and
Course Technology, among others.
For more information, please visit http://oreilly.com/safari.
How to Contact Us
Please address comments and questions concerning this book to the publisher:
O’Reilly Media, Inc.
1005 Gravenstein Highway North
Sebastopol, CA 95472
800-998-9938 (in the United States or Canada)
707-829-0515 (international or local)
707-829-0104 (fax)
Appendix A.
Thank You!
Before we close the last chapter of this book, I would like to thank you for reading it
up to the last paragraph. I truly hope that you had as much pleasure reading this book
as I had writing it, and that it will be useful for your projects, big or small.
If you find errors, please send feedback. More generally, I would love to know what
you think, so please don’t hesitate to contact me via O’Reilly, or through the ageron/
handson-ml GitHub project.
March 10, 2026, 11:09 a.m.
Q: What is gradient descent and how does it work?
# Gradient Descent
Based on the context provided:
**What it is:** Gradient Descent is a very generic optimization algorithm capable of finding optimal solutions to a wide range of problems.
**How it works:** The general idea of Gradient Descent is to tweak parameters iteratively in order to minimize a cost function.
**The process:** The algorithm computes the gradient of the cost function with regards to each parameter in the network, then uses these gradients to update each parameter with a Gradient Descent step.
**Efficiency note:** The context illustrates that feature scaling matters for performance—with proper scaling, Gradient Descent can go straight toward the minimum and reach it quickly. Without it, the algorithm may first go in a direction almost orthogonal to the global minimum and take a long time traversing a flat valley before reaching the minimum.
3 source chunks
Figure 4-7. Gradient Descent with and without feature scaling
Gradient Descent | 113
Download from finelybook www.finelybook.com
As you can see, on the left the Gradient Descent algorithm goes straight toward the
minimum, thereby reaching it quickly, whereas on the righ
t it first goes in a direction
almost orthogonal to the direction of the global minimum, and it ends with a long
march down an almost flat valley. It will eventually reach the minimum, but it will
take a long time.
110 | Chapter 4: Training Models
Download from finelybook www.finelybook.com
Gradient Descent
Gra
dient Descent is a very generic optimization algorithm capable of finding optimal
solutions to a wide range of problems. The general idea of Gradient Descent is to
tweak parameters iteratively in order to minimize a cost function.
Suppose you are lost in the mountains in a dense fog; you can only feel the slope of
the output layer to the input layer, propagating the error gradient on the way. Once
the algorithm has computed the gradient of the cost function with regards to each
275
Download from finelybook www.finelybook.com
1 “Understanding the Difficulty of Training Deep Feedforward Neural Networks, ” X. Glorot, Y Bengio (2010).
parameter in the network, it uses these gradients to update each parameter with a
Gradient Descent step
.
March 10, 2026, 11:09 a.m.
Q: What are the main types of machine learning?
Based on the context provided, Machine Learning systems can be classified into broad categories based on two main dimensions:
1. **Whether or not they are trained with human supervision:**
- Supervised Learning
- Unsupervised Learning
- Semisupervised Learning
- Reinforcement Learning
2. **Whether or not they can learn incrementally on the fly:**
- Online Learning
- Batch Learning
The context also mentions that ML systems can be further classified in other ways, such as instance-based or model-based, though specific details about these categories are not provided in the given text.
3 source chunks
To summarize, Machine Learning is grea
t for:
• Problems for which existing solutions require a lot of hand-tuning or long lists of
rules: one Machine Learning algorithm can often simplif
y code and perform bet‐
ter.
• Complex problems for which there is no good solution at all using a traditional
approach: the best Machine Learning techniques can find a solution.
•
Fluctuating environments: a Machine Learning system can adapt to new data.
• Getting insights about complex problems and large amounts of data.
Types of Machine Learning Systems
There are so many different types of M
achine Learning systems that it is useful to
classify them in broad categories based on:
• Whether or not they are trained with human supervision (supervised, unsuper‐
vised, semisupervised, and Reinforcement Learning)
• Whether or not they can learn incremen
tally on the fly (online versus batch
learning)
big picture:
28 | Chapter 1: The Machine Learning Landscape
Download from finelybook www.finelybook.com
• Machine Learning is about making machines get better at some task by learning
from data, instead of having to explicitly code rules.
• There are man
y different types of ML systems: supervised or not, batch or online,
instance-based or model-based, and so on.
• In a ML project you gather data in a training set, and you feed the training set to
March 10, 2026, 11 a.m.
Q: What is scikit-learn used for?
Based on the context provided, Scikit-Learn is used for **Machine Learning**. Specifically, it is described as "very easy to use" and "implements many Machine Learning algorithms."
The context indicates it is a production-ready Python framework used to implement various Machine Learning techniques, ranging from simple methods like linear regression to more complex Deep Learning techniques. It provides classes and APIs for training models such as Decision Tree classifiers, as well as ensemble methods like bagging and pasting.
3 source chunks
CPU cores or even differen t ser
vers. Similarly, predictions can be made in parallel.
This is one of the reasons why bagging and pasting are such popular methods: they
scale very well.
Bagging and Pasting in Scikit-Learn
Scikit-Learn offers a simple API for both bagging and pasting with the BaggingClas
sifier class (or BaggingRegressor for regression). The following code trains an
ensemble of 500 Decision Tree classifiers,5 each trained on 100 training instances ran‐
on all samples at each node. This results in a training complexity of O(n × m log(m)).
For small training sets (less than a few thousand instances), Scikit-Learn can speed up
training by presorting the data (set presort=True), but this slows down training con‐
siderably for larger training sets.
Gini Impurity or Entropy?
By default, the Gini impurity measure is used, but you can select the entropy impurity
measure instead by setting the criterion hyperparameter to "entropy". The concept
ment programs capable of learning from data.
We will cover a large number of techniques, from the simplest and most commonly
used (such as linear regression) to some of the Deep Learning techniques that regu‐
larly win competitions.
Rather than implementing our own toy versions of each algorithm, we will be using
actual production-ready Python frameworks:
• Scikit-Learn is ver y easy to use, yet it im plemen
ts many Machine Learning algo‐
March 10, 2026, 11 a.m.