As a Machine Learning Engineer, I remember feeling uncertain about the path ahead when I first started out. Back then, I had no idea that the title “Machine Learning Engineer” even existed when I was working on my Text-to-Speech study for my Bachelor’s thesis. But it has become a popular and sought-after title, particularly among computer science students. With this in mind, I created a roadmap specifically for those interested in pursuing a career in this field. The roadmap starts with the basics and gradually builds up skills, one layer at a time. It’s worth noting that in today’s world and job market, having a Master’s degree is almost a requirement to land a job in this sector, as it can make a tremendous difference.
What are the prerequisites to becoming a Machine Learning Engineer?
It all starts with math. Linear Algebra, Statistics and Probability are our first requirements before jumping into ML algorithms. It’s important to have a strong foundation in these areas since Machine Learning is based on statistical and probabilistic models.
- Books
- Naked Statistics is a good book to start
- Mathematics for Machine Learning
- Videos
- Khan Academy
- MIT
Programming is the second skill set before jumping into ML algorithms. Machine learning involves a lot of mathematical and statistical concepts, so it’s essential to have a good understanding of data structures and algorithms.
- Google has good libraries to develop your programming skills.
- Cracking Coding Interview is a great book to develop your programming skills.
- Intro to Python Course: it can be any language/course but Python is a well-known programming language to practice ML.
- Develop a basic SQL understanding. An introductory video as follows: https://www.youtube.com/watch?v=h0nxCDiD-zg&ab_channel=KevinStratvert
- You need to practice to be able to develop your programming skills so Hackerrank and Leetcode are great for this.
You’re trusting your math-related skills, then we can start to talk about Machine Learning, and it all starts with basic Machine Learning Algorithms. Moreover, you can start to build up skills in libraries like Pandas, scikit-learn, and Numpy at this point.
- Books
- Python Data Science Handbook
- Approaching Almost Any Machine Learning Problem
- Videos
- Coursera’s Supervised Machine Learning
- Kaggle’s Intro to Machine Learning
- Blogs
- Google’s Machine Learning Collection
- Zero to Mastery ML
- Also, from now on, Kaggle and Huggingface are where you’re going to spend time building a deeper understanding and practising what you learn.
a- If your path lies in NLP
- I’m a massive fan of Spacy, and I think you should consider learning it from the following playlist.
- Speech and Language Processing by Jurafsky and Martin
- If you’d like to keep a book, then Natural Language Processing with Python is good.
- Kaggle’s learning platform for NLP
- Huggingface’s learning platform for NLP
- Must-read papers: NLP is extensive, and I personally couldn’t limit it to 10 papers; however, I highly encourage you to read the articles in the following link.
b- If your path lies in Computer Vision, first, you must program using OpenCV. It’s a vision library that’s easy to learn and use.
- Introduction to Computer Vision
- Deep Learning for Computer Vision
- Computer Vision Lectures
- I like the must-read research papers section for computer vision in the following blog post.
Finally, let’s not forget about deep learning algorithms. While it can be combined with item 4, I prefer to keep it separate to emphasize its significance in modern machine-learning applications.
- I think Misra Turp did a fantastic job with the following video series on YouTube. It is a great place to start learning: 50 Days of Deep Learning
- Andrew Ng’s Deep Learning Specialization
- Deep Mind and UCL also released a video series on Youtube: Deep Learning Lecture Series 2020
- Ian Goodfellow’s Deep Learning book
- Zero to Mastery Deep Learning using TensorFlow
While learning is essential, it is also necessary to put that knowledge into practice and showcase your abilities to increase your chances of securing a job. With that in mind, I would strongly advise participating in a competition on Kaggle and taking a task by joining Huggingface’s Discord.
There is one thing that you should continue even after you land a job following the latest research in your field. And believe it or not, this is not easy to keep up with the literature. There are a couple of places I personally used to follow:
- Papers with Code
- ArXiv
- I also benefit from Huggingface’s blog and task section: https://huggingface.co/blog?tag=research
- Another resource is from CohereAI for me to follow: Best NLP Research Papers in 2022
Thanks for Reading!
If you enjoyed this, follow me to don't miss another article on data science guides, tricks and tips, life lessons, and more!
Connect with me
LinkedIn : https://www.linkedin.com/in/durgesh4993/
GitHub : https://github.com/Durgesh4993
LeetCode : https://leetcode.com/durgesh4993/
Top comments (0)