Today Machine Learning is the most Buzzed word in whole Tech Industry. Everyone talk about it. But most of them do not find the right path , how to learn Machine Learning. Here I tried to simplify the path.
Lets get to the point, Where to start and How to start : I will try to answer both questions together using Important Links Step by Step.
1. Mathematics for Machine Learning:
You can start Machine by introducing some mathematical concepts. These are building blocks of many Machine Learning Algorithms.
- Linear Algebra ( You have to deal with various matrix operations, dimension reduction techniques , higher dimension visualization etc). There is a course on MIT 18.06 Linear Algebra by Prof. Gilbert Strang. There is another fun course on Linear Algebra by 3Blue1Brown.
- Probability and Statistics ( Machine Learning and Statistics are not very different fields. Basically, Machine Learning is “Statistics doing on Computers”. Introduction to Probability — The Science of Uncertainty by MITx on Edx.
- Calculus (Calculus has many roles in Machine Learning but the most important is Optimization of cost function.) Essence of calculus by 3Blue1Brown.
- Others ( Algorithm & Complexity, Real and Complex Analysis, Information Theory, etc.)
2. Learn Programming Language :
But for the starting point of view. I would recommend, You should learn first Python (which is more of Industry Focus) and R (which is more of Research Oriented.) Introduction to Computer Science and Programming Using Python by MITx.
Along with that try to learn their many libraries such as Pandas, Numpy, Matplotlib, etc. You can learn this by codebasics on Youtube.
3. Introduction to Machine Learning :
Now you can start learning ML. In my Opinion, you should start with a course provided by Andrew Ng Machine Learning. As the assignments are in Octave, try to implement in Python or in R.
- If You want more mathematical intuition take Learning from Data: A Short Course by Yaser S. Abu-Mostafa.
- You can also take Intro to Machine Learning by Udacity.
- Along with that try to learn most popular ML framework Scikit-learn.
After Finishing above courses, You can go more Deeper in Machine Learning using Deep Learning.
4. Deep Learning :
- Start with a Deep Learning Specialization by Andrew Ng. This consist of 5 wonderful courses (Deeplearning.ai)
- Neural Networks and Deep Learning.
- Improving Deep Neural Networks: Hyperparameter tuning, Regularization and Optimization.
- Structuring Machine Learning Projects.
- Convolutional Neural Networks. Sequence Models.
- Learn Few Deep Learning Framworks such as Tensorflow, Keras, Pytorch, etc.
- If you want more practical course take Fast.ai by Jeremy Howard.
Have A Happy Learning…
Source: Deep Learning on Medium