Factorized Decision Trees for Active Learning in Recommender Systems.
Rasoul KarimiMartin WistubaAlexandros NanopoulosLars Schmidt-ThiemePublished in: ICTAI (2013)
Keyphrases
- recommender systems
- active learning
- decision trees
- matrix factorization
- training set
- collaborative filtering
- machine learning
- decision tree induction
- information overload
- random sampling
- predictive accuracy
- learning strategies
- naive bayes
- training data
- learning algorithm
- product recommendation
- machine learning algorithms
- imbalanced data classification
- cold start problem
- user profiles
- decision tree learning
- experimental design
- trust aware
- decision rules
- supervised learning
- user preferences
- user modeling
- constructive induction
- data sparsity
- user profiling
- random forest
- semi supervised
- pool based active learning
- transfer learning
- music emotion classification
- feature construction
- recommendation quality
- active learning strategies
- personalized recommendation
- classification rules
- data sets
- selective sampling
- labeled data
- training examples
- user model
- data mining methods
- version space
- information gain
- cold start
- batch mode
- rule induction
- implicit feedback
- sample complexity
- sample selection
- classification accuracy
- rating prediction
- active learning framework
- decision tree algorithm