How to start learning Machine Learning?
After giving a lot of thought to whether you want to be a machine learning engineer or not, you decided to become one. Now I am not going to say that if you decide anything to do then half of your work is already done, no that is so not true. There are lots of things to learn in machine learning, and you are probably going to get distracted, confused, lost, and maybe something more. Cause it happened with me, with my friends, and mostly with all the other machine learning engineers. I know, I know we learn more from our failures/mistakes, not from others, but if you did the same mistake everybody else did then you are no different from anybody else. So be smart and learn something from other's mistakes.
Before knowing how to learn ML, let’s know about its history:
ML is a study of computer algorithms that improve automatically through experience. The term Machine Learning was coined in 1959 by Arthur Samuel, an American IBMer. He was a pioneer in the field of computer gaming and Artificial Intelligence.
One of the formal and widely used definitions of ML is, “A computer program said to be learned from experience E with respect to some class of tasks T and performance measure P. If it’s a performance at tasks in T, as measured by P, improves with experience E.” provided by Tom M. Mitchell.
If the definition doesn’t make any sense don’t worry, you will understand it later.
Buzz Words:
When it comes to ML you will come across many different buzz words such as Artificial Intelligence, Neural Network, Deep Learning. Now let’s try to understand what do they mean and how they are related to each other.
- Artificial Intelligence: AI refers to the simulation of human intelligence in machines that are programmed to think like humans and mimic their actions, in simple worlds archive human intelligence to a computer.
- Neural Network: A construct in Machine Learning inspired by the network of neurons (nerve cells) in the biological brain. Neural networks are a fundamental part of deep learning.
- Deep Learning: A subfield of machine learning that uses multi-layered neural networks. Often, “machine learning” and “deep learning” are used interchangeably.
Note: Machine learning is a subfield of Artificial Intelligence.
Key points:
- Mathematics
- Programming Language
- Machine Learning Concepts
- Machine Learning Practice
1. Mathematics:
Before learning something there are always Prerequisites and maths in one of the Prerequisites to learn ML, but how much maths do we need to know? Do we have to be experts in maths? What if I am very weak in maths? This type of question starts depressing us.
No, you don’t have to be an expert in math. If you have basic knowledge about Linear Algebra, Multivariate Calculus, and Statistics then you are good to go. Some people prefer to skip learning maths and learn them as they go, so it’s ok if you don’t know that much about Linear Algebra, Multivariate Calculus, and Statistics. You just need to get up and start learning.
2. Programming Language:
Now, everyone mostly confuses themselves while learning a programming language. Either they can’t decide which one to choose, or they just use all the languages to learn ML. If you think that learning ML in all the possible languages is the best idea, then at some point you will get confused between all syntaxes, you will west most of your precious time coding in a different language, but what you need to do is clear ML concept, and practice more ML models. So just choose only one language and go with it.
You can learn ML in Python, R, Scala, and many more. Which one to chooses? How to decide which language will be best for you? Now if you ask me, I will suggest Python, all because of the different types of libraries, which makes your task so simple, and also python is very easy to learn, But in the end, which one to choose is totally depends on which type of ML engineer you want to become. If you want to be a Statistician, Analyst, or Data Scientist then you can choose R, but if you want to be a Data Engineer or Data Scientist then you can choose Python.
Now, how much deep do we have to learn Programming Language? You just need to know all the basics about the language, basic functions, object-oriented concepts, a few data structures, and algorithms. And now you are good to go.
3. Machine Learning Concepts:
Now that you have Mathematical and programming knowledge, then it’s time to start learning ML(Which is the fun Part!!!)
There are lots of ways to get started with ML, you can either enroll in any online programs/courses, you can use Books, or you can read online documents/blogs.
- Online Courses for ML:
- Machine Learning by Andrew Ng
- Machine Learning crash course by Google developers
- Machine Learning A-Z: Hands-on Python & R by Udemy
2. Books for ML:
- Machine Learning For Absolute Beginners: A Plain English Introduction (2nd Edition) by Oliver Theobald
- Machine Learning (in Python and R) For Dummies (1st Edition)
3. Online Blogs/documents:
4. Machine Learning Practice:
Now here comes the question in your mind, ok so far I know enough to get started with ML practice, so how do I get started?
Well, everyone has their own ways, I started practicing the ML model while I was learning it from Machine Learning A-Z: Hands-on Python & R by Udemy, after that I got to know about Kaggle, Kaggle was developed by Google in 2010, it has a lot of datasets and it also contains courses where you can practice, create you own models, and discuss your problems with different peoples.
You can start practicing on HackerEarth also, it contains tutorials and problems to practice.
And the best way to start learn something is to try it by yourself, so start building your own model.
Don’t stop after this, there is a lot to learn and you just started it, so best of luck…