I am from Bangladesh and i think my country needs AI badly.Is there anyone who can give a view on how i should approach?
Nuhash Jnr
Posted on
Can Anyone Guide Me How To be good at Artificial Intelligence and Machine Learning?
For further actions, you may consider blocking this person and/or reporting abuse
Top comments (10)
Hey Nuhash. I think to start AI and ML you should probably start by reading stats and then referring to any other material would make sense. Also, after you are done with stats, linear algebra and calculus you can refer to the Hands on Machine learning using scikit learn and tensorflow book followed by some practice on Kaggle. That would pretty much make you strong and confident enough to help your country using your skills and tools.
Have a nice day. Appreciate your effort.
I was looking for such replies :)
Thanx a lot Saswat.Appreciate it
It's one thing to follow the tutorials and learning how to use the development tools, and another to understand the essence of machine learning models and how and why they work. Regarding machine learning you need understanding in three fields:
It may also help to join Society for Industrial and Applied Mathematics (SIAM). From Bangladesh, you are eligible for the Outreach Membership.
If you like to learn in a top-bottom approach, check machinelearningmastery.com/start-h...
A good top-to-bottom approach would be fast.ai by Jeremy Howard. I think this is a good way to start since you would build things right away, before going deep into theory.
Once you get your feet wet with Machine Learning, I suggest taking Andrew Ng's Stanford Machine Learning Course in Coursera. This is the gold standard on machine learning courses! The only problem here is that it uses the Octave programming language (opensource alternative to MATLAB), which is considerably dated. However, there are already a lot of Python implementations of his exercises in the internet, all you have to do is a quick search. I cannot stress enough the wealth of information that you can get from this course! The pros certainly outweigh the cons. Note that the math is a little bit involved, so I suggest you brush up on your calculus (or learn them as you see them).
Lastly, I suggest learning about statistics, since this is what AI is all about under the hood. Think Stats and Think Bayes (available at greenteapress.com) are good books for programmers to learn statistics. It follows the same style as Jeremy Howard: learn by doing.
I think a lot of people will disagree with this approach of starting from the top before learning the theory and the details, but here's why I think it's effective. First, it allows you to build things from the get go, so you will feel progress right from the start. This increases your motivation so you keep on going. Second, it lets you understand quickly what you don't know. This let's you understand the gaps that you need to fill once you start learning the theory. And lastly, it makes you feel like a freaking boss being able to create AI models right from the start 😎
Good luck on your journey, brother
Thanks Brother for your opinion
A good start would be to dive into the machine learning tutorials at RealPython
Okay, I will take a look. Thanx JIM :)
I am also a web developer not a machine learning expert so I cannot suggest any guide. But I have taken this Coursera Machine Learning by Andrew Ng which I believe is one of the most valuable and comprehensive courses I have ever taken online. And it is free!
Good luck!
Follow @juliasilge here on dev.to :)