Ivy Mar 30, 2016 No Comments
We had written about one of our students Shan who won a Hackathon Contest. He shared with us his experiences as a Data Scientist and what is his “Mantra”of success.
Excerpts from Interview:
“Congratulations on your win! How do you feel to win it among 2500+ contestants?
Shan: Thank you. Its really inspiring competing with top data scientists across the globe. In this type of competitive endeavor the learning curve is steep. The experiences and inspirations will motivate me in future endeavors.
What inspired you to participate in such contests? Since when have you been participating?
Shan: Any hackathon requires a unique blend of skills and different factors. In this, persistence is very important. It is easy to become demotivated when you are not placed in the top slots. Its definitely worth it keep trying. This persistence needs to come from within. One should ask why he/she is engaging in a particular competition. Learning is the motivator? Getting opportunities by ranking high on the leaderboard? Or, just want to prove himself/herself? The monetary reward in most competitions is not enough to motivate a significant time investment, so unless you clearly know what you want and how to motivate yourself, it can be tough to keep trying.
I started off learning about interesting things that may or may not impact score. Trying out of box ideas and viewing its impact on score. Then gradually build up inspiration when I frequently featured in Top25 for few competitions.
For last one year , I have been participating in AnalyticsVidhya, KDnuggets , kaggle and different forums.
Do you think it helps?
Shan: It helps. Here, you can work on industry data. Whatever concepts we have acquired, we can put them in action. Trying out different ways to approach a problem .
What skills are needed to win such contests?
Shan : I can explain them as follows :
1.Most important is analytical skill. Developing a hypothesis statement keeping objective in perspective can give a break through. It saves lot of time.
2.Next is Data Exploration and Data preparation. Graphical representations can give interesting insights. Feature Engineering improves scores. Algorithm selection is important, but in my opinion Feature Engineering plays a great role. Insights from graphs can help to develop interesting features.
3.Domain-specific knowledge can be helpful, but it is not critical, can generally be picked up by learning while one is competing. One need to quickly learn the relevant domain-specific elements that he/she don’t know
4.It’s more interesting to do something new and learn about it as you go along. Machine Learning algorithms help getting goods scores. Most of the contests are won using ML algos
5.Persistence plays a major role. Converging on a good model takes time, so to keep going, it really helps to be enthusiastic about the problem, or about learning something new,
What is your suggestion to budding Data Scientists?
Shan: Winning a contest is an agglomeration of lot of small insights. Keep learning, keep working to iterate your solution and have confidence you will do well.'”
Thanks Shan for your time. Keep Winning and keep sharing your Mantras for success!
All the Best
Team Ivy
Leave a Reply