Land your first job in Machine Learning: Part 1
This blog is a summary of the book: Introduction to Machine Learning Interviews Book by Chip Huyen.
Photo by Sincerely Media on Unsplash
Chip Huyen is a writer,computer Scientist and the co-founder of Claypot AI, a platform for real-time machine learning.
She has worked in companies like Google, NVIDIA, Snap, Netflix, Primer AI, and Snorkel AI. She has worked as an executive in the hiring process and she has also made a curriculum for many people to land their first Machine Learning job.
In this book, she has given some tips and tricks to help you get the job in the ML industry.
First of all in order to get a job in Machine Learning Industry you need to know the kind of role you want to get into.
There is a great video on Youtube that describes all the aspects of Machine Learning Engineering What machine learning role is right for you? by Josh Tobin, Full Stack Deep Learning Bootcamp 2019.
Some of them are described here below :
Machine Learning in Research
Finding solutions to basic issues and adding to the amount of theoretical knowledge are the two objectives of research. Research projects usually involve employing scientific procedures to verify if a hypothesis or theory is correct.
Machine Learning in Production
The goal of production is to create or enhance a product. A product can be a good (e.g. a car), a service (e.g. ride-sharing service), a process (e.g. detecting whether a transaction is fraudulent), or a business insight.
Generally, companies use a bit of both approaches, but if you want to pursue chose the latter because there are more roles involving production than roles involving research.
The table below is taken from the article Stanford’s CS 329S, lecture 1: Understanding machine learning production.
Research | Production | |
Objectives | Model performance | Different stakeholders have different objectives |
Computational priority | Fast training, high throughput | Fast inference, low latency |
Data | Static | Constantly shifting |
Fairness | Good to have (sadly) | Important |
Interpretability | Good to have | Important |