I bought my first fitbit recently and I am loving it!
While tracking my activities is obviously good, what makes it a great product is the eco-system it provides. Every day, I am competing with my friends to finish those extra steps and shed those extra kilos! We all challenge each other and buck up each other (sometimes in not so polite way as well) to make the group a fitter group!
How is a Fitbit tracker related to Massive Open Online Courses (MOOCs)?
Well, staying fit without an eco-system and learning new tools by yourself have a few common challenges:
- Both of them are very easy to start, but difficult to sustain and accomplish
- You need to have Super-mannish willpower to think you can master / accomplish them by yourself.
So, in this post, I am sharing some best practices, I have learnt the hard way. I have undergone more than 50 open courses on Data Science across various platforms (including Coursera, edX, Udacity, SAS and several others) in last 18 months or so. I still sign up for most of the new courses which are floated.
1. Decide what is there for you in that course?
One of the things, people get wrong is that they assume that entire course is for everyone to be undergone in same way. This is certainly not true. It is true that the creators of the courses try hard to make the courses as generic for the larger audience, but you still need to decide and define what you are looking for in each of these courses.
For example a course on machine learning would typically start by whetting your appetite about machine learning. If I was undergoing that course, I don’t need that section. So I’ll typically skip that section (P.S. It may be a good idea to do the exercises – more on this later).
Before you start the course, you should have following things defined:
- What all sections would you cover?
- How much time will you spend per week? How much more or how less it is compared to the expected time?
- What is the overall objective from the course? For example, by undergoing Learning from Data on edX, I wanted to have an overview of all commonly used machine learning algorithms and be in top 20% of a Kaggle competition.
2. Create an eco-system – find mentors and buddies
This is where I think Fitbit does a fab job! You need to almost replicate it for MOOCs. Here is how I typically find them.
Mentors: Typically people in your network on Linkedin, experts in the fields, bloggers on the subject. Just reach out to them and ask for help as a mentor. I usually define the expectations from them up front as well – how much time I expect from them? what kind of questions / projects I would be working upon? I also try and offer them some thing in return – it could be research / tutorial for the blogger, a presentation for the expert presenting in a conference – you get the idea!
Buddies: Best is to reach out to people who are asking questions similar to the ones you have. Again, key is to interact with them before finalizing. You would want to have some one who is determined to finish the course, participating in projects and trying out things outside the templates in the MOOC. A good buddy can make or break your learnings from a MOOC. Once found, you should spend time discussing your understanding and doubts with these buddies.
Another practice, which can be of help in some places is to join meet-ups and find people with similar interest.
3. Participate in discussions
Discussions are probably one of the best way to learn while experimenting. There is a reason why every coder in this world spends time on stackoverflow to get his answers. We run a similar discussion portal for data science professionals here. All the MOOCs will typically have their own discussion portals as well. If they are active, you should fire your questions there as well.
While on the disucssion portal, offer what ever help you can to fellow participants.
4. Blog about your learning by applying them to various other problems
Another way to showcase your learning, is to blog about them. This is how I started my journey! Once you have learnt about a particular technique – apply it to another dataset and come up with a solution. Then see what other people have done. iPython notebooks and Github can help immensely here. You just need a functional blog, where you can note what you are doing, what you are doing and how you are doing.
5. Participate in projects, assignments and competitions
Projects and assignments are lifeline of all the data science MOOCs. If you miss on them, you miss on a lot of learning. Again, different platforms have different approach. Udacity, particularly is more project focused than others. So, make it a point to do all the assignments diligently – even if you think they are too easy / basic or they are too tough. Just have a go at it.
If you don’t add few numbers of an array or a list thinking that it is too easy, you will struggle with the data cleaning later in a project. If you can afford, you should also participate in a few competitions outside of the MOOC using the same tools you have learnt. So, if you have learn Logistic regression, apply that on Kaggle Titanic survivor competition.
6. Stay on the schedule
There are 2 commitments you need to make in order to make sure you don’t drop out during the process. The first of them is to stay on the schedule. Finish the content and assignments on every week with in the week itself. Don’t play the catch up game, otherwise you will fall out the next week.
This might sound easier that what is actually is. You will need to put in a lot of efforts to make sure that you are on the top of the schedule.
7. Decide the minimum benchmark, you will achieve irrespective
The second commitment you should make to yourself is the minimum score you would achieve. This is a personal choice and should be in sync with the need of the course. If this is an important course, you should at least commit 70% score.
End Notes:
I hope this post would help numerous people who struggle to complete open courses. They are a fabulous platform to learn new skills. If you have any other tips and tricks to help MOOC learning, please feel free to share
On the lighter side, if you use a Fitbit to track your fitness – let’s compete!
If you like what you just read & want to continue your analytics learning, subscribe to our emails, follow us on twitter or like our facebook page.
Kunal is a post graduate from IIT Bombay in Aerospace Engineering. He has spent more than 10 years in field of Data Science. His work experience ranges from mature markets like UK to a developing market like India. During this period he has lead teams of various sizes and has worked on various tools like SAS, SPSS, Qlikview, R, Python and Matlab.
