How to Build Better Supervised Models? Learnings from the Industry and Data Science competitions

Auditorium

This talk will be around building better models and covering the topics like:

 

  • Build Better Models
  • Steps in a Data science problem solving
  • Why is domain knowledge required to do well in Data science problem-solving?
  • Be Paranoid about Generalization
  • Focus on Last Mile Optimization (Debug & Never give up)
  • Case Study of Winning Data Science Competition: ACM KDD
  • Problem Background
  • We should also talk about learnings as part of the talk – what worked and what did not work
  • My Winning Solution Components
Machine Learning
Social media & sharing icons powered by UltimatelySocial