Facebook V: Predicting Check Ins

The article on my blog explains the details of the approach.

High level instructions to go from the raw data to the winning submission are mentioned in the "Instructions" pdf. Please feel free to discuss anything that might be unclear in the topic of the winning submission: https://www.kaggle.com/c/facebook-v-predicting-check-ins/forums/t/22081/1st-place-winning-solution.

Also make sure to check out my exploratory analysis Shiny application: https://tvdwiele.shinyapps.io/Facebook-V/

Name		Name	Last commit message	Last commit date
Latest commit History 20 Commits
Candidate selection		Candidate selection
Common		Common
Data		Data
Downsampling		Downsampling
Evaluate predictions		Evaluate predictions
Exploratory analysis		Exploratory analysis
Feature engineering		Feature engineering
First level learners		First level learners
References		References
Second level learners		Second level learners
Strategy		Strategy
Submission		Submission
.gitattributes		.gitattributes
.gitignore		.gitignore
Instructions.docx		Instructions.docx
Instructions.pdf		Instructions.pdf
README.md		README.md
winningSubmission.7z		winningSubmission.7z

Provide feedback