Kaggle competition to personalize hotel search Learn from the historical data what types of hotels are bought/clicked Rank those hotels higher. Results evaluated by the common search ranking evaluation - normalized discounted cumulative gain The dataset can be obtained at http://www.kaggle.com/c/expedia-personalized-sort/data
The "code" folder contains the following:
- household - module that manages matrices of datapoints in processing
- ndcg - calculates normalized discounted gain for internal validation
- processAllTrain - given data from Kaggle - this normalizes the data and add features
- softKmeanHotel - implement Kmean on hotel
- svm implement svm on hotel
- svmRank writes data in the form that works with SVM rank http://www.cs.cornell.edu/people/tj/svm_light/svm_rank.html
- svmRankEval evaluates using ndcg from the predicted result