credits : pratik-ratadiya
”If I had 100 days to build a basic proficiency in ML, what path would you suggest?” ❓
A common beginner’s block - so I decided to document a 5-step roadmap for 100 days of ML learning! ⚡1. Do some introductory coursework (5 weeks): Hands-on ML is easy but not unique unless you understand the working of these algorithms. A good theory course is helpful. I recommend one of:
a. Machine Learning Specialization by Andrew Ng (Coursera)
b. Stanford CS229 (YouTube)
c. Machine learning crash course (Google Developers)
Or in books: An Introduction to Statistical Learning (ISLR)
2. Learn Hands-on ML in Python (3 weeks): The best way is through practice, but boot camps can help cover the essentials. I recommend one of:
a. Python for Data science and ML bootcamp (Udemy)
b. Machine learning A-Z - Hands-on Python & R (Udemy)
Or in books: Hands-on ML with Sklearn, Keras, and Tensorflow (O'Reilly)
3. Review how to do EDA (1 week): Exploratory Data Analysis and Feature engineering are crucial blocks in the ML pipeline. Explore tutorials and existing work on Kaggle and GitHub. I am adding two examples in the comments.
4. Take part in your first Kaggle contests / Do your first two mini-projects (4 weeks): Start implementing your skills through some competitions on Kaggle: Titanic, House prices, and Digit recognizer are good starting sets.
Alternatively, how about doing two mini-projects?: Build your models using some datasets from the UCI Machine Learning Repository.
5. Read and plan for the future (1 week):
By now you have basic proficiency in working with ML algorithms. ML is a broad field, and you should now start thinking about sub-domains to explore (CV, NLP, RL, Optimization, Security, Fairness, etc.). Give time to read articles, work, and experiences of people in these fields. Plan your next steps.
This 100 days of preparation should give you 2 mini-projects, 1 rigorous theoretical coursework, and at least 6 weeks of hands-on experience. A strong foundation in ML. Pass it on!
Machine Learning Specialization: https://www.coursera.org/specializations/machine-learning-introduction
Stanford CS229: https://www.youtube.com/playlist?list=PLoROMvodv4rMiGQp3WXShtMGgzqpfVfbU
Machine learning crash course: https://developers.google.com/machine-learning/crash-course/
ISLR: https://www.statlearning.com/
Python for Data science and ML Bootcamp: https://www.udemy.com/course/python-for-data-science-and-machine-learning-bootcamp/
Machine learning A-Z - Hands-on Python & R: https://www.udemy.com/course/machinelearning/
Hands-on ML by O’Reilly: https://www.oreilly.com/library/view/hands-on-machine-learning/9781492032632/
EDA on Titanic dataset: https://www.kaggle.com/code/ash316/eda-to-prediction-dietanic
EDA on text (Comment toxicity): https://www.kaggle.com/code/jagangupta/stop-the-s-toxic-comments-eda
MACHINE LEARNING IN APPLE INC
The question I have been asked the most since my Apple offer: "How do I prepare for Machine learning interviews?"
Here are 7 free Machine learning study resources to get you ready for that dream interview:
1. ML Concepts: Introduction to Statistical Learning
A must read, this book covers all the standard ML algorithms in detail. A good mix of theory, mathematics, and diagrams. It also provides implementation labs in R.
Book link: https://lnkd.in/gfNtNabg
2. ML Mathematics: Mathematics for Machine Learning
Not everyone will work on them, but if you want proficiency in mathematical concepts that govern ML algorithms (often needed for research), this book is a good read.
Book link: https://lnkd.in/g-W2vf8i
3. Deep learning: The Deep Learning Book
Stalwarts in the domain: Ian Goodfellow, Yoshua Bengio, and Aaron Courville come together to write this book - suited for both students and professionals. Covers everything deep learning: theory to practice!
Book link: https://lnkd.in/gA9dF5bB
4. Domain-specific concepts: Colah's Blog
Chris Olah has some of the best notes and walkthroughs on various neural networks. Just the perfect way for you to quickly brush up on a particular architecture type.
Blog link: https://colah.github.io
5. Interview Questions: Introduction to ML Interviews
Prof. Chip Huyen does an amazing job in compiling not only 200+ ML interview questions, but also providing a detailed overview of the ML interview process. Must read for everyone.
Book link: https://lnkd.in/gpEFxHsh
6. MLOps + Production: Made with ML
This course by Goku Mohandas
covers every aspect of the Machine Learning pipeline. Being aware of these practices and tools will give you a huge advantage over others.
Course link: https://madewithml.com/
7. Problem solving: Leetcode Patterns
Need to dedicate time for problem solving as well. Sean Prashad has made a great compilation of problems from Leetcode that covers specific companies and patterns. Can't find a quicker way to refresh those concepts!
Problem List: https://lnkd.in/gaFR7PRK
8. Your resume!
Goes without saying, but make sure you know every aspect of those ML projects you have on your resume. Your motivation behind choosing an algorithm, metrics, datasets, and advantages - everything.
So, these are the resources I used for my interview prep. I recommend them not only because they are free, but because they are as quality a learning material as you can possibly get. Now to go on and master the domain!
_________________________________________________________________________________
𝗠𝗮𝗰𝗵𝗶𝗻𝗲 𝗟𝗲𝗮𝗿𝗻𝗶𝗻𝗴
1-Introduction to Machine Learning
🌀https://lnkd.in/ecqatZBA
2-Stanford CS229: Machine Learning
🌀https://lnkd.in/exNpHVgK
3-Making Friends with Machine Learning
🌀https://lnkd.in/ejM83n2B
4-Applied Machine Learning
🌀https://lnkd.in/eBPxJriR
5-Statistical Machine Learning
🌀https://lnkd.in/eszaHhnn
6-Neural Networks: Zero to Hero
🌀https://lnkd.in/eHyXfdMY
𝗗𝗲𝗲𝗽 𝗟𝗲𝗮𝗿𝗻𝗶𝗻𝗴
1-Introduction to Deep Learning (MIT)
🌀https://lnkd.in/eh7wTzbq
2-CMU Introduction to Deep Learning
🌀https://lnkd.in/eCF92CDW
3-MIT: Deep Learning for Art, Aesthetics, and Creativity
🌀https://lnkd.in/eURJKHUS
4-Stanford Deep Learning
🌀https://lnkd.in/e6fWv3R8
5-Introduction to Deep Learning (MIT)
🌀https://lnkd.in/eh7wTzbq
6-CMU Introduction to Deep Learning
🌀https://lnkd.in/eCF92CDW
7-Deep Unsupervised Learning
🌀https://lnkd.in/eB2sacxY
8-NYU Deep Learning SP21
🌀https://lnkd.in/eMNsFmBe
9-Foundation Models
🌀https://lnkd.in/emPfE-MS
10-Full Stack Deep Learning
🌀https://lnkd.in/edbFQZBX
11-Practical Deep Learning for Coders
🌀https://lnkd.in/eMjrFESY
12-Machine Learning Engineering for Production (MLOps)
🌀https://lnkd.in/enmaT8Yy
𝗔𝗿𝘁𝗶𝗳𝗶𝗰𝗶𝗮𝗹 𝗜𝗻𝘁𝗲𝗹𝗹𝗶𝗴𝗲𝗻𝗰𝗲
1-NLP Course (Hugging Face)
🌀https://lnkd.in/ehhG4S_2
2-Natural Language Understanding
🌀https://lnkd.in/emVV8CfA
3-CMU Advanced NLP 2022
🌀https://lnkd.in/eXenh9pm
4-Multilingual NLP
🌀https://lnkd.in/eu46qy8V
5-Advanced NLP
🌀https://lnkd.in/ggpQAD6
1-Deep Learning for Computer Vision
🌀https://lnkd.in/eMk3kWSz
1-Deep Reinforcement Learning
🌀https://lnkd.in/e6gyvp4s
2-Stanford: Reinforcement Learning
🌀https://lnkd.in/eGR-5THW