A bit of Background
It was late 2017 when I took up a book on Machine Learning(ML) from a well known publisher. I went about reading couple of chapters and before I start the next, I was already demotivated, tired and started having a feeling, 'Not My Cup of Tea'. Well, there was nothing wrong about the book or the content. It was how I approached it without considering the limitations I had. Also, educating myself on ML was not my top listed learning priority.
Having experience in working on Web Technologies for long time, the next cool thing I wanted to do was mobile app development. Once that craze over, Machine Learning was the next. With an average expertise on Engineering Mathematics and Statistics, I got the mental push back and had to re-plan my approach for learning Machine Learning.
The New Approach
I took an alternate approach towards my Learning. This blog is about that Journey.
Do a Course at my speed
With the initial set back, I wanted to go with something that I would be comfortable with. Early 2018, I got to know about this ever green resource on Coursera: Andrew Ng's Course on Coursera.
It took me almost 14 weeks to complete the course. But anyone with full dedication to it, can complete the course in just 3 weeks time with an hour per day of involvement. This course was an eye opener to me. Extremely engaging, simple explanation and lots of hand holding helped me to get all my interest back for the Subject. You can also get a certificate after completion of the course, in case it motivates further.
Stay Connected to know ML
While Andrew Ng's Course was taking its sweet time, I wanted to stay engaged to the subject. I can not thank enough to this Audio Podcast by OCDevel. This is an excellent resource to go through in parallel with anything else that you are doing on Machine Learning. OCDevel's podcast was(and is) a companion in my Home-Office-Home Car Drive.
Next, few best things that happen to me was,
- Jason Brownlee's Newsletter from Machine Learning Mastery . Once subscribed, I started getting emails explaining concepts, sample chapters and lots of knowledge links on ML.
- Getting Involved with Analytics Vidya Community. You can find the App on Google Store
A Piece of Motivation
All that mentioned above were not just as one-time-references. I tried visiting all these references time to time whenever I had the need.
My notebook is my biggest motivation ๐๐๐. Often it gets dusted. I clean it, read it and feel good about the journey I have made so far!
Programming Languages and tool
As it goes with most of the programming concepts, we need the support of a Programming Language to realize the concept better. Here are few that I have been learning for last few months:
- Python (I am in Love with it)
- Pandas
- TensorFlow
What Next
Next 3-6 months, I would like to continue to write small - medium size programs on various aspects of Learning, Prediction and Error. I am also in plan to explore Deep Learning and conceptualize it better.
This was originally posted on my HashNode Blog.
Top comments (3)
Thanks for this post.
Sir I'm a full stack web developer and wants to learn AI and ML
. i wanna ask you one thing how to learn math topics for ML and AI please help me for this
Thanks shahzaibanwar009 for liking the post.
I was also in the same shoe as yours. The Coursera course I have mentioned in the post takes care of Maths as well. It speaks about the Linear Algebra, Statistics in details with lots of practices and quizzes. I would highly recommend the course at your pace.
At times, I needed some extra information when a concept was introduced. For example, I am good at Vectors, Matrix etc but couldn't understand Gradient Decent Concept initially. I took additional help by searching it in youtube and then again going back to the course.
You might find you own ways though. Best of luck. Feel free to share, how is it going.
Thanks for help