a data mining course using a kaggle competition · i started the competition in the final month...
TRANSCRIPT
![Page 1: A DATA MINING COURSE USING A KAGGLE COMPETITION · I Started the competition in the final month and had some integration with the class (displayed the leaderboard at the beginning](https://reader035.vdocuments.mx/reader035/viewer/2022071218/60527d0999a5c52dc2455b15/html5/thumbnails/1.jpg)
. . . . . .
A DATA MINING COURSEUSING A KAGGLE
COMPETITIONSusan Holmes, Nelson Ray ©
April 24, 2011
ABabcdfghiejkl
![Page 2: A DATA MINING COURSE USING A KAGGLE COMPETITION · I Started the competition in the final month and had some integration with the class (displayed the leaderboard at the beginning](https://reader035.vdocuments.mx/reader035/viewer/2022071218/60527d0999a5c52dc2455b15/html5/thumbnails/2.jpg)
. . . . . .
HOW TO GET STARTED IN DATAMINING?
If you have plenty of time:
I Learn statisticsI Learn RI Learn linear algebraI Learn functional
analysis (kernels,..)I Learn data base
management.
![Page 3: A DATA MINING COURSE USING A KAGGLE COMPETITION · I Started the competition in the final month and had some integration with the class (displayed the leaderboard at the beginning](https://reader035.vdocuments.mx/reader035/viewer/2022071218/60527d0999a5c52dc2455b15/html5/thumbnails/3.jpg)
. . . . . .
HOW TO GET STARTED IN DATAMINING?
If you have plenty of time:
I Learn statisticsI Learn RI Learn linear algebraI Learn functional
analysis (kernels,..)I Learn data base
management.
![Page 4: A DATA MINING COURSE USING A KAGGLE COMPETITION · I Started the competition in the final month and had some integration with the class (displayed the leaderboard at the beginning](https://reader035.vdocuments.mx/reader035/viewer/2022071218/60527d0999a5c52dc2455b15/html5/thumbnails/4.jpg)
. . . . . .
BEST NEURAL NETWORKAVAILABLE
You have it:
![Page 5: A DATA MINING COURSE USING A KAGGLE COMPETITION · I Started the competition in the final month and had some integration with the class (displayed the leaderboard at the beginning](https://reader035.vdocuments.mx/reader035/viewer/2022071218/60527d0999a5c52dc2455b15/html5/thumbnails/5.jpg)
. . . . . .
BEST NEURAL NETWORKAVAILABLE
You have it:
![Page 6: A DATA MINING COURSE USING A KAGGLE COMPETITION · I Started the competition in the final month and had some integration with the class (displayed the leaderboard at the beginning](https://reader035.vdocuments.mx/reader035/viewer/2022071218/60527d0999a5c52dc2455b15/html5/thumbnails/6.jpg)
. . . . . .
BEST NEURAL NETWORKAVAILABLE
You have it:
![Page 7: A DATA MINING COURSE USING A KAGGLE COMPETITION · I Started the competition in the final month and had some integration with the class (displayed the leaderboard at the beginning](https://reader035.vdocuments.mx/reader035/viewer/2022071218/60527d0999a5c52dc2455b15/html5/thumbnails/7.jpg)
. . . . . .
DATAMINING IN ONE COURSE
We need a hands on approach.
![Page 8: A DATA MINING COURSE USING A KAGGLE COMPETITION · I Started the competition in the final month and had some integration with the class (displayed the leaderboard at the beginning](https://reader035.vdocuments.mx/reader035/viewer/2022071218/60527d0999a5c52dc2455b15/html5/thumbnails/8.jpg)
. . . . . .
HOW TO GET REALLY STARTED ?
![Page 9: A DATA MINING COURSE USING A KAGGLE COMPETITION · I Started the competition in the final month and had some integration with the class (displayed the leaderboard at the beginning](https://reader035.vdocuments.mx/reader035/viewer/2022071218/60527d0999a5c52dc2455b15/html5/thumbnails/9.jpg)
. . . . . .
HOW TO GET REALLY STARTED ?
![Page 10: A DATA MINING COURSE USING A KAGGLE COMPETITION · I Started the competition in the final month and had some integration with the class (displayed the leaderboard at the beginning](https://reader035.vdocuments.mx/reader035/viewer/2022071218/60527d0999a5c52dc2455b15/html5/thumbnails/10.jpg)
. . . . . .
AVOID BLACK BOXES
![Page 11: A DATA MINING COURSE USING A KAGGLE COMPETITION · I Started the competition in the final month and had some integration with the class (displayed the leaderboard at the beginning](https://reader035.vdocuments.mx/reader035/viewer/2022071218/60527d0999a5c52dc2455b15/html5/thumbnails/11.jpg)
. . . . . .
LINEAR ALGEBRA: THE IGON VALUE PROBLEM
Malcolm Gladwell made an unfortunate error in`What the dog saw' (in his spelling of eigenvalue).Igon Value Problem is one of dilettantism.We have to understand the methods and not treat them asblack boxes.
![Page 12: A DATA MINING COURSE USING A KAGGLE COMPETITION · I Started the competition in the final month and had some integration with the class (displayed the leaderboard at the beginning](https://reader035.vdocuments.mx/reader035/viewer/2022071218/60527d0999a5c52dc2455b15/html5/thumbnails/12.jpg)
. . . . . .
START WITH DATA YOU KNOW WELL
I Because you invented it (simulation studies).I It is your domain of expertise.I You have already studied it with other tools.
![Page 13: A DATA MINING COURSE USING A KAGGLE COMPETITION · I Started the competition in the final month and had some integration with the class (displayed the leaderboard at the beginning](https://reader035.vdocuments.mx/reader035/viewer/2022071218/60527d0999a5c52dc2455b15/html5/thumbnails/13.jpg)
. . . . . .
START IN THE SUPERVISED CONTEXT
Because there is a gold standard for evaluation.
![Page 14: A DATA MINING COURSE USING A KAGGLE COMPETITION · I Started the competition in the final month and had some integration with the class (displayed the leaderboard at the beginning](https://reader035.vdocuments.mx/reader035/viewer/2022071218/60527d0999a5c52dc2455b15/html5/thumbnails/14.jpg)
. . . . . .
FINISH OFF BY DOING A REAL COMPETITION
Data Mining competitions with Kaggle.Started at the middle of the course, students who ranked intop 3 didn't have to take the final.
![Page 15: A DATA MINING COURSE USING A KAGGLE COMPETITION · I Started the competition in the final month and had some integration with the class (displayed the leaderboard at the beginning](https://reader035.vdocuments.mx/reader035/viewer/2022071218/60527d0999a5c52dc2455b15/html5/thumbnails/15.jpg)
. . . . . .
PICK AN INTERESTING DATASET
![Page 16: A DATA MINING COURSE USING A KAGGLE COMPETITION · I Started the competition in the final month and had some integration with the class (displayed the leaderboard at the beginning](https://reader035.vdocuments.mx/reader035/viewer/2022071218/60527d0999a5c52dc2455b15/html5/thumbnails/16.jpg)
. . . . . .
WINE HAS MANY GREAT STORIES!
I Robert Parker is arguably themost influential wine critic.
I ``The Robert Parker Effect'':wines ranked 90 points costsignificantly more than thoseranked 89 points.
I Each spring since 1994 RobertParker evaluates many wines inBordeaux.
I In 2003, at the request of hisfamily, he didn't make the trip.
I Economists have estimated thatthis depressed wine prices byabout 15%.
![Page 17: A DATA MINING COURSE USING A KAGGLE COMPETITION · I Started the competition in the final month and had some integration with the class (displayed the leaderboard at the beginning](https://reader035.vdocuments.mx/reader035/viewer/2022071218/60527d0999a5c52dc2455b15/html5/thumbnails/17.jpg)
. . . . . .
PLENTY OF STRUCTURED INFORMATION
![Page 18: A DATA MINING COURSE USING A KAGGLE COMPETITION · I Started the competition in the final month and had some integration with the class (displayed the leaderboard at the beginning](https://reader035.vdocuments.mx/reader035/viewer/2022071218/60527d0999a5c52dc2455b15/html5/thumbnails/18.jpg)
. . . . . .
IN ONE TIDY PACKAGE
![Page 19: A DATA MINING COURSE USING A KAGGLE COMPETITION · I Started the competition in the final month and had some integration with the class (displayed the leaderboard at the beginning](https://reader035.vdocuments.mx/reader035/viewer/2022071218/60527d0999a5c52dc2455b15/html5/thumbnails/19.jpg)
. . . . . .
KAGGLE-IN-CLASS
![Page 20: A DATA MINING COURSE USING A KAGGLE COMPETITION · I Started the competition in the final month and had some integration with the class (displayed the leaderboard at the beginning](https://reader035.vdocuments.mx/reader035/viewer/2022071218/60527d0999a5c52dc2455b15/html5/thumbnails/20.jpg)
. . . . . .
THE COMPETITION PAGE
![Page 21: A DATA MINING COURSE USING A KAGGLE COMPETITION · I Started the competition in the final month and had some integration with the class (displayed the leaderboard at the beginning](https://reader035.vdocuments.mx/reader035/viewer/2022071218/60527d0999a5c52dc2455b15/html5/thumbnails/21.jpg)
. . . . . .
THE RESULTS
![Page 22: A DATA MINING COURSE USING A KAGGLE COMPETITION · I Started the competition in the final month and had some integration with the class (displayed the leaderboard at the beginning](https://reader035.vdocuments.mx/reader035/viewer/2022071218/60527d0999a5c52dc2455b15/html5/thumbnails/22.jpg)
. . . . . .
THE WINNERS REVEAL THEIR STRATEGY
![Page 23: A DATA MINING COURSE USING A KAGGLE COMPETITION · I Started the competition in the final month and had some integration with the class (displayed the leaderboard at the beginning](https://reader035.vdocuments.mx/reader035/viewer/2022071218/60527d0999a5c52dc2455b15/html5/thumbnails/23.jpg)
. . . . . .
POST-MORTEM
I At least 20 teams from Stats 202 participated.I Teams were limited to three people.I Roughly 160 students in the class.I Contest ran for about a month with a one submission per
day limit.I The top 3 teams averaged more than one submission
every other day.I Received many thank yous from students for the
competition.
![Page 24: A DATA MINING COURSE USING A KAGGLE COMPETITION · I Started the competition in the final month and had some integration with the class (displayed the leaderboard at the beginning](https://reader035.vdocuments.mx/reader035/viewer/2022071218/60527d0999a5c52dc2455b15/html5/thumbnails/24.jpg)
. . . . . .
IDEAS FOR IMPROVEMENT
I Started the competition in the final month and had someintegration with the class (displayed the leaderboard atthe beginning of class and had some homeworkassignments involving the data)
I Could start even earlier and provide sample code usingincreasingly sophisticated techniques (i.e. starting withlinear models and moving to basic text mining forfeatures in boosted trees).
I Code would be provided at regular intervals to allow newparticipants to easily catch up and push the leaders evenfurther.
I Have ``progress prizes'' so the top teams don't entirelyneglect studying for the final.