Join us on 5/19 at the H2O offices for Wendy Kan of Kaggle to talk about how she designs Kaggle contents followed by a talk from Mark Landry, H2O on how he has sat at the very top for most of "how much does it rain" challenge.
Schedule for the evening:
7:00 - 7:20 :: Mingling & Munching
7:20 - 7:40 :: Wendy Kan, Kaggle: How she designs Kaggle contests plus q&a
7:40 - 7:50 :: Anthony Goldbloom: demo their new Kaggle Scripts
7:50 - 9:00 :: Mark Landry, H2O: How much does it rain?
9:00 - 9:30 :: Mingling
Kaggle is a popular global data science competition platform that has more than 300,000 data scientists in the community. Each competition attracts hundreds to thousands of participants to compete building the
best predictive model. As the data scientist, my job involves
interacting with competition hosts, understanding their data and business needs, and help them answer questions by designing data science competitions based on their data, so the crowd can help them achieve their goals. In this talk, I will share my experience designing Kaggle competitions.
Wendy Kan, Data Scientist
Before joining Kaggle, Wendy worked as a researcher and software engineer at GE Research and Genentech, developing predictive models/analytical tools for pharmaceutical research. She holds bachelor's and master's degrees in Electrical Engineering from National Tsing Hua University, and PhD in Biomedical Engineering from
University of Texas at Austin.
Mark Landry is an experienced Kaggler (prize winner and seven-time Top 10% competitor), and will share his pro tips on using H2O for his current challenge, How Much Did It Rain? Mark will discuss:
- an overview of the data
- how to do data prep work efficiently using data.table in R
- GBM modeling in H2O
- model output analysis (how solutions to different parts of the same problem lead to very different models)
Join us and learn from a Kaggle pro how H2O can help you develop a challenge-winning solution!
Mark Landry is a competition data scientist and product manager at H2O. He enjoys testing ideas in Kaggle competitions, where he is ranked in the top 100 in the world (top 0.03%) and well-trained in getting quick solutions to iterate over. Most at home in SQL, he found H2O through hacking in R. Interests are multi-model architectures and helping the world make fewer models that perform worse than the mean.