This is my solution for the Kaggle competition - Facebook Recruiting IV: Human or Robot?
Facebook Recruiting IV: Human or Robot?
There are two datasets in this competition. One is a bidder dataset that includes a list of bidder information, including their id, payment account, and address. The other is a bid dataset that includes 7.6 million bids on different auctions. The datasets can be downloaded from
The solution is done in three seperate R scripts
-
Analysis.R
-
Analysis_testing.R
-
Classification.R
The following are the steps done
- Merge the train and the bid datasets by bidder_id to create the training dataset
- Check what could be the possible features
- Create the Tidy training and testing datasets based on the chosen features
- Choose the appropraite classification algorithm based on the ROC test metric
- Make the appropraite prediction of the outcome of the testing dataset
You can find additional information about the variables, data and transformations in the CodeBook.md file.