6 answers
1
Question
Asked
487 views
I want to become a Machine learning Engineer, where should i start my journey?
#engineer #computer #technology
Login to comment
6 answers
Updated
Dr. Shaul’s Answer
Dear Sashank. The term ML Engineer is somewhat ambiguous. If you want to work as a data scientist you should get some theoretical understanding of this area and most importantly start doing ML coding.
For a basic intro my favorite book is "The Hundred-Page Machine Learning Book" - http://themlbook.com/. There are also numerous online courses, just google for "ML for beginners" and give them a try. One I recommend is https://www.kaggle.com/learn/intro-to-machine-learning. I am not a big fan of courses that emphasize math and linear algebra, but if you are, try Andrew Ng's famous Coursera (Stanford) course - https://www.coursera.org/learn/machine-learning.
Once you have some confidence try a couple of Kaggle competitions - see link at the end of the Kaggle course. E.g. housing prices for regression and the Titanic for classification.
If by ML Engineer you mean more focus on data try to learn about big data and scalable ML applications (this is what I teach alongside my work :-), and do some coding with Kafka and Spark including Spark ML. Udemy's "The Ultimate Hands-On Hadoop: Tame your Big Data!" is a simple and nice intro in my opinion.
Good Luck!
-- Shaul
For a basic intro my favorite book is "The Hundred-Page Machine Learning Book" - http://themlbook.com/. There are also numerous online courses, just google for "ML for beginners" and give them a try. One I recommend is https://www.kaggle.com/learn/intro-to-machine-learning. I am not a big fan of courses that emphasize math and linear algebra, but if you are, try Andrew Ng's famous Coursera (Stanford) course - https://www.coursera.org/learn/machine-learning.
Once you have some confidence try a couple of Kaggle competitions - see link at the end of the Kaggle course. E.g. housing prices for regression and the Titanic for classification.
If by ML Engineer you mean more focus on data try to learn about big data and scalable ML applications (this is what I teach alongside my work :-), and do some coding with Kafka and Spark including Spark ML. Udemy's "The Ultimate Hands-On Hadoop: Tame your Big Data!" is a simple and nice intro in my opinion.
Good Luck!
-- Shaul
Updated
Alex’s Answer
Hello Sashank! I am currently a Product Manager for a software development team, and share your interest in machine learning and advanced technology. Artificial intelligence is a very exciting space that is disrupting almost every industry, and machine learning engineers are in high demand. To get started, I would recommend 1) reading intro-level books on machine learning that cover the basic concepts and real world use cases and 2) learning more about Python, which has become the most prevalent development language for ML engineers.
Updated
Bryan’s Answer
Hi Shashank,
I recommend the following to mimic an intro to machine learning course. It will require lots of self-learning and covers the fundamentals for machine learning.
Develop a strong foundation in Python
Read "Intro to Machine Learning using Python" https://github.com/dlsucomet/MLResources/blob/master/books/%5BML%5D%20Introduction%20to%20Machine%20Learning%20with%20Python%20(2017).pdf
I recommend the following to mimic an intro to machine learning course. It will require lots of self-learning and covers the fundamentals for machine learning.
Bryan recommends the following next steps:
Updated
D S’s Answer
Start playing with data using Python. Check for relevant courses on Coursera and Udemy.
Updated
Sameer’s Answer
First you can do some simpler things like getting familiar with the tools and technology. So try readymade examples using simple tools like Jupyter notebook and Python. That will give you an idea of what Machine Learning can actually do.
Then pick an area say NLP and do some formal training via online learning portals like Udemy, Corsera etc.
Now if you feel you need to go deeper and understand more from mathematical or statistical point of view, then do a course by Andrew NG.
If you realize that you like playing with Data then learning Big Data, data quality management, creating Warehouses and then applying machine learning techniques on top of that would be another good path.
Then pick an area say NLP and do some formal training via online learning portals like Udemy, Corsera etc.
Now if you feel you need to go deeper and understand more from mathematical or statistical point of view, then do a course by Andrew NG.
If you realize that you like playing with Data then learning Big Data, data quality management, creating Warehouses and then applying machine learning techniques on top of that would be another good path.
Updated
Dominic’s Answer
As Dr. Shaul mentioned, someone who uses ML and AI to solve business problems is a Data Scientist.
In my opinion, to be good at data science you need to have some knowledge in statistics and in programming (preferably in Python or R), and a passion to learn.
As a data scientist, I find that the more I learn about ML, the more I discover that I know less. I am currently in Computer Science program with a concentration in Machine Learning at Georgia Tech, and I feel like I am always learning a lot from taking the courses.
While formal education can be very useful, I know a lot of colleagues who have become incredibly great data scientists through self-learning.
Back to your question on how you can get started... I think the best way is to engage in a fun personal/work project, a hackathon, or a competition on Kaggle.com. This will frustrate you for a while but when you get your project to work, your eyes will brighten up in gratification and it will motivate you to continue learning more and get better. Examples of personal projects you can work on include predicting stock/bitcoin prices, scraping tweets and predicting the sentiments from users on certain topics, predicting sports outcomes for your fantasy football/basketball team etc. A lot of the datasets that you may need are available online for free.
Another advantage of using Kaggle is the sense of community. You can engage with like-minded people and learn from each other. Also, try to follow data science/ML groups on social media platforms so that you stay engaged and even become an active contributor over time.
All the best Sashank!
-Dom
In my opinion, to be good at data science you need to have some knowledge in statistics and in programming (preferably in Python or R), and a passion to learn.
As a data scientist, I find that the more I learn about ML, the more I discover that I know less. I am currently in Computer Science program with a concentration in Machine Learning at Georgia Tech, and I feel like I am always learning a lot from taking the courses.
While formal education can be very useful, I know a lot of colleagues who have become incredibly great data scientists through self-learning.
Back to your question on how you can get started... I think the best way is to engage in a fun personal/work project, a hackathon, or a competition on Kaggle.com. This will frustrate you for a while but when you get your project to work, your eyes will brighten up in gratification and it will motivate you to continue learning more and get better. Examples of personal projects you can work on include predicting stock/bitcoin prices, scraping tweets and predicting the sentiments from users on certain topics, predicting sports outcomes for your fantasy football/basketball team etc. A lot of the datasets that you may need are available online for free.
Another advantage of using Kaggle is the sense of community. You can engage with like-minded people and learn from each other. Also, try to follow data science/ML groups on social media platforms so that you stay engaged and even become an active contributor over time.
All the best Sashank!
-Dom