I always found Kaggle competition quite fascinating because the problems are interesting and various a lot. We can always learn some new tricks and gain some valuable insights from each competition. Since most of the comps are real-life cases, it is also quite practical and meaningful to handle them well. Since I found participating in each comp and getting a gold medal can usually take quite long time (>100 hours). I decide to actually step back and learn from those comps by reading through the discussion and notebook after the comp has finished. Let’s see how much we can gain from this journey. Hope we will find something valuable through this series as well.
The main things I wanna know from each competition is:
- What’s special about the data?
- How to understand the data better?
- What model works?
- What competition tricks work?