Competitive, prize money driven, science dates back to the wager in the 1800s to disprove the flat-earth hypothesis. Then there was the wager between Tesla, the other Tesla, and Thomas Edison to improve DC-current power generation. Sadly, the loser refused to pay up on both those occasions. In modern times that at least is not a concern. This talk is about Data Science Competitions with special reference to the Kaggle Data Science web-portal. In the first half, Shantanu describe some past competitions organized by Kaggle, the standardized structure of the competitions, how to take part and the resources available to support you.
In the second half, the speaker will run us through a competition he is currently taking part in. It’s the impossible task of forecasting the outcomes of March Madness during this year's NCAA Division I Men’s and Women’s Basketball Championships. He is currently trending in the Top 20s in the leaderboard for the competitions. Shantanu will discuss hardware, software, techniques used, tips and tricks for planning and organization, the benefits of competing and of course managing the emotional roller-coaster ride than any competitive activity entails.
About our Speaker
Shantanu Karve, a member of our own Westlake Village Data Science Meetup has over 15 years of experience as Lead Data Scientist at Valen Analytics, Farmers Insurance, and Legalzoom. He is also an active Kaggler where he consistently breaks into the top of the leader-board on various competitions including $1M Zillow’s challenge to improve home value prediction as well as DHS passenger screening challenge to improve the accuracy of threat recognition.
Prior to his career in data science he spent many years in senior IT roles at ICI, Mars, and as the UK Product Manager for the then 2nd largest computing company in the world, Digital Equipment Corporation. Shantanu holds a degree in mathematics from Bath University in the UK.