expedia-kaggle

• Using pandas to import traning data
• Extract datetime, year and month from training data
• Select a random sample of 10000 records out of 37 million records
• Select 2013 and early 2014 and mark it t1
• Select late 2014 and mark it t2
• Remove non-booking data
• Do PCA on destination
• Extact Checkin and Checkout dates
• RandomForest from sklearn
• Group by search destination distance
• Obtain most frequent hotels