There are thousands of machine learning algorithms out there.

But nobody has time to learn them all - And you probably don't need to.

Instead, learn just these 6 machine learning algorithms (that you will probably use 90% of the time in your career): Regression

The hello world of machine learning.

This is a simple algorithm to learn and:

• Serves as a baseline
• Is easy to interpret
• Can be used for online settings

Don't scoff at linear/logistic regression. Next we have...

Machine learning interviews are hard.

Preparing for them can be stressful - There's way too much to learn.

Here are 5 simple steps you can use to crack these interviews (And resources you can use to land your ML dream job)👇 Coding Skills (DSA)

Most people think grinding leetcode is the way

Wrong! Think in patterns: Just 100-150 problems are enough

If you do this, you will:

• Identify solution paths faster
• Avoid burnout
• Learn principles vs memorize code

Resource: leetcode.com

Time and Attention are our most valuable resources

What's the point of wasting them learning things over and over - only to forget them again?

3 simple tricks that will help you remember what you learn 50x better: Blind Summary

Shortly after you've read/watched something, write it in your own words without looking at the source.

This will:

• Check your understanding
• Uncover blind spots
• Force you to review

I use this all the time on new concepts.

But this next tip ...

Jupyter notebooks are awesome - If only you could manipulate them faster.

Well, you can. I have combed through nearly every Jupyter tutorial on the Internet.

Many tips are interesting. But only few are worth your time.

Here are my 5 favorites—that will transform your workflow Accidental Cell Deletions 🤦🏼‍♂️

Deleted a cell in Jupyter by mistake? Ctrl/Cmd + Z not doing the job?

Use Esc + Z

Thank me later. 🤌🏼

Training a model is just half the battle.

An accurate model is useless if you can't deploy it -- So make it fast and small.

These 3 tips will make your model fast & lightweight. Knowledge Distillation (Jedi technique)

Your model is like Obi Wan. Take a smaller model (Padawan Luke Skywalker) and train it using:

• Ground truth labels (The force)
• Teacher's labels (Obi Wan)

You will end up with a smaller model that can be as good as the teacher.

I spent over 1,000 hours learning machine learning effectively— because I started late in my ML journey.

Unfortunately, there's no universal recipe for everyone.

Here are 5 things I've tried (that you can start applying today to launch your ML journey): Targeted Practice

List out skills you want to develop— Then, practice those skills hands on.

For example:

• Practice data cleaning on datasets
• Play directly with libraries like Sklearn
• Interpreting training curves
• Tuning existing models

This focuses your learning.

5 must-have machine learning skills everyone should know (that most courses and schools don't teach you): 🧼 Dataset Cleaning and Curation

Learning to build your own dataset is invaluable.

Focus on :

• Formatting & Labeling
• Removing Outliers & Duplicates
• Diversity and Bias Reduction

A great dataset makes models sing. There is no machine learning without data.

Debugging Neural Networks 101

Over the last 2 years, I have trained 1000s of models successfully.

Here's a secret - I use the same 5 steps every time.

My debugging time has come down from hours to minutes (And now yours can too) 🔶 Remove the Fancy Stuff

Inherited a model? To debug errors, eliminate fluff.

So remove:

• Mixed Precision.
• Regularization like dropout
• Early Stopping, Schedulers, etc.

This allows you to laser focus on model errors.

Teach me Machine learning without teaching me machine learning in 3 minutes.

Challenge Accepted 🧵👇🏼 Many years ago in Spain, there were two best friends, Ben and Will.

Ben wanted to be a painter. But he sucked.

Will's dream was to be a curator. He was an assistant at the local museum.

Would they ever achieve their goals?

What is the fastest way to get good at Machine Learning? 🚀

Meticulously studying other's code.

Here are 5 of the best git repositories to significantly accelerate your growth in ML 🧵

#ML #Python #MachineLearning Vision Transformers

Phil Wang, a.k.a lucidrains has one of the best transformer implementations out there. The code is easy to read, and beautifully structured.

If attention is all you need, lucidrains code is all you need:

github.com/lucidrains/vit…

The Classical Machine Learning Roadmap. 🗺️

(Read this if you want to get the next best thing to a degree in machine learning without spending $$$$$, stop running through an endless cave of resources and build a career in ML)

#MachineLearning #ML #DataScience 🧵👇 Each of us is built differently - Some like theory; Some like Code. I'll show a pathway for both

What you need:
- Curious mindset
- Ability to Code

Criteria for resource inclusion:
- No Deep Learning*
- Free or very affordable
- High ROI

Where do we start?

The First-Time Machine Learning Playbook.

(Read this if you want to efficiently learn machine learning, avoid frustration from searching for resources, and build a career in ML.)

#Ship30For30 #ML #MachineLearning What people think you need:

Most people think you need to be a math and stats expert to learn ML. ML can be math intensive, but it’s not a barrier to entry.

What you need:

• The Ability to code
• An Open and Curious Mind
• A Good starting point

Which leads me to …