66 Days of Data: It’s not about perfection

Round 1

What is #66daysofdata?

#66daysofdata is an initiative started by Ken Jee. The tagline is

66 days. 5 minutes a day. Every day.

While learning Data Science, we get too overwhelmed by the vastness and extraordinary things it has to give us. I felt this thing right after I learnt about few algorithms, Deep Learning, MLops and more. It’s been more than a year now that I started learning about the Machine Learning field and so on. As Ken Jee’s #66daysofdata site says:

Learning data science is about consistency over time. 66Days of Data is about establishing great data science habits around daily practice and sharing your work.

All one has to do is study Data Science for 66 days regularly. The minimum criteria met should be 5 minutes per day and tweet on Twitter or LinkedIn about your journey. If you want to know more about it, watch this:

https://www.youtube.com/watch?v=qV_AlRwhI3I&t=3s

Why can’t I be consistent like others?

I started #66daysofdata two times, but I couldn’t study it every day and apply it. I was regular for a maximum of about 11/15 days, and then I couldn’t follow it up more. When I was again starting #66daysofdata for the third time, I sat down and realised the mistakes that I was making the first two times, and I came down to these things:

1. I didn’t plan what I wanted to achieve throughout these 66 spans of days.
2. I directly jumped to different topics instead of studying the earlier ones and get a good grip on the basic concepts.
3. The obvious one, Procrastinated a lot.
4. I focused on too many things at once — Exploratory Data Analysis, Deep Learning, Research Paper, etc. There was too much on the table.

Adrenaline Rush

It was time that I complete the challenge now and omit this procrastination. I made a list of the things that I want to achieve through these 66 days. The list felt too overwhelming again, and I was on the verge of giving up again. Once I completed the 15 days, my adrenaline kick that now I’ve reached the wall of 15 days, and now it is easily achievable. There were days when I felt like I don’t want to study, but I somehow managed to do something for 5 minutes.

After completing the 66 days, I was ready to move to the Round 2 of 66 days of data and not just get this done! I took a few days to break and started with #66daysofdata Round 2. I’m currently on my 2nd Round, and the adrenaline kick is already high every day.

Have a look here at my #66daysofdata Round 1 journey:

https://twitter.com/rohan_kumawat/status/1368633483870892032?s=20

Discovery and Review

Here’s a list of all the things I learned and reviewed again:

1. Kaggle Competitions (Tabular Playground Series — March, April, May).
2. Linear and Logistic Regression.
3. Label encoding, One Hot Encoding.
4. K-Nearest Neighbours.
5. Decision tree.
6. Boosting Algorithms.
7. Hyper Parameter Optimization.
8. Cross-validation techniques.
9. Dimensionality Reduction.
10. Artificial Neural Network.
11. Data Mining and Data Warehousing.
12. Image Matching.
13. Principal Component Analysis.
14. Autonomous Vehicles.

I wasn’t able to complete the Kaggle Competitions successfully. I made notes on all of these things and build up my concept. This Round 1 wasn’t perfect, and if I got to rate my performance out of 10, I would rate 3/10, but it was worth pursuing. If I hadn’t been consistent for 66 days, then I don’t think that I would have cleared my concepts on a more conceptual level.

Round 2 Adrenaline Kick

I’ve made a list of what I want to do in Round 2, and one of the most important things I want to do is “Show my Work”. So, this time I’ll come up with blogs on all of the things that I’ve learnt so far and the things I’m learning nowadays. I’m in the middle of Round 2 right now, which is more fun than Round 1. I’m not looking for perfection here but continuous throughout the 66 days with a motto of “Show my Work”.

I encourage you to do this and join me and the others who are in the middle of this journey. If you are starting or are in the middle of your Data Science journey, please DM me over Twitter or LinkedIn for your accountability go-to guy.