Santiago Profile picture
May 6 6 tweets 3 min read
There are 20,001 machine learning and data science books out there.

Great, but everyone needs something different.

We created @0xbnomial, and we wrote a book to teach you something new, one question at a time:

bnomial.gumroad.com/l/2204

But there's something even better: ↓
The most frequently asked question since we launched @0xbnomial:

Where can I find old questions?

This is the place!
All 30 questions we published in April are here.

We revisited and expanded every explanation and updated each reference.

On top of that, we tell you what the EASY, MEDIUM and HARD questions are because we have the data!
Here is the "something better" you are looking for:

If you buy the book today, you'll need to buy it next month and the next one.

A much better deal:

Subscribe, get every book as soon as published, and get access to every paid feature:

bnomial.gumroad.com/l/bnomial
I hope you enjoy this book.

But most importantly, I hope you show up every day and answer @0xbnomial's question.

It's free. It's fast. It's fun (FFF)

Don't forget the Power of Tiny Gains!
So far, this was a brilliant idea.

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with Santiago

Santiago Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @svpino

Apr 29
Beginners think that all it takes to recognize objects in an image is to download the latest favorite pre-trained model.

I've been asked, "why would that be hard?" more than once.

Here is what you are missing.
You need to consider how the data is captured.

Is it fixed cameras? Is the hardware standard? Are illumination conditions the same across the board?

Are all pictures taken from the same angle?

Is the quality of the images roughly the same?
You need to consider the frequency of new images coming in.

How much do you need to scale to process new images?

How do you store the images? How about security and privacy?

How fast will data drift impact your model? How frequently do you need to retrain it?
Read 7 tweets
Apr 28
The most exciting machine learning book of 2022.

I've been waiting for it for quite some time and finally, Amazon dropped it last Sunday.

Link: amzn.to/3K57Wdn

If you are starting with machine learning, here is why you want to read "The Kaggle Book": ↓
Last week, I mentioned—and I quote:

"You don't know shit until you try things in the real world. Read all you want. It doesn't matter if you don't practice it."

Do you know the perfect way to put everything you learn into practice?

Join Kaggle!
As somebody who wants to start with machine learning, Kaggle is *THE* place to be.

But up until today, information about Kaggle has been scattered in blog posts and online conversations.

"The Kaggle Book" is here to close that gap.

And this book is definitely for you!
Read 10 tweets
Apr 26
Multi-class classification is not the same as multi-label classification.

Yet many people starting with machine learning don't understand the difference.

Let's solve that problem once and for all.

1 of 6
We have many pictures of animals.

We want to classify them into three different classes based on the animal: a cat, a dog, or a chick.

(In our hypothetical world, there are only these three animal species.)

2 of 6
This problem, as I just presented it to you, is a multi-class classification problem:

• We have samples (images)
• Each sample belongs to a class

We assume that a picture is either a cat, a dog, or a chick, but not a combination of them!

3 of 6
Read 6 tweets
Apr 25
A great opportunity!

After 8 weeks, students from the School of AI will present the work they did to solve different challenges.

And you can attend online for free!

Register here → pischool.link/sv

Here is why you shouldn't miss this: ↓
First, @picampusschool runs the School of AI: an 8-week AI program where you learn about:

• NLP, Speech Recognition
• Time Series Forecasting
• 3D Face Animations
• Startup and Business Forecasting

18 of you got a free scholarship!
Now, you can register to see the result of this session on May 6th.

This is one of a kind chance to hear about solutions to real-life challenges.

You learn new techniques, and ideas, and see the hurdles that students had to face to get here.

And that's only the beginning.
Read 4 tweets
Apr 22
You are building a binary classification model to determine whether a picture shows a hot dog or not.

Which of the following activation functions is a good candidate for the output layer?

1. Rectifier Linear Unit (ReLU)
2. Leaky ReLU
3. Sigmoid
4. Softmax

Think about it.
The answer:

You could use either Sigmoid or Softmax.

I'm sure most of you picked Sigmoid. It's certainly the best way to approach this problem.

But Softmax is also an option —although unconventional for a binary classifier.

How about the other choices?
ReLU returns its input if positive, zero otherwise. Leaky ReLU is based on ReLU but with a slight slope for negative values.

ReLU and Leaky ReLU will always give us the previous layer output with no modifications if the input is always positive.

This doesn't help us.
Read 10 tweets
Apr 19
Writing machine learning questions is not simple, especially when publishing one question every day.

Here is the process I follow to write questions that —hopefully– help you learn something new.

I currently follow a 4-step process:

1. Write the essence/core of the question
2. Write a story around it
3. Peer review it
4. Do one round of edits before publishing it
The first step is to decide what exactly I want to ask.

I start with the core question and its four possible answers.

Nothing else: no story, no explanation, no references.

This is about the essence of the question.
Read 8 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us on Twitter!

:(