Risk-Sensitive Learning to Rank with Evolutionary Multi-Objective Feature Selection.
Daniel Xavier de SousaSérgio D. CanutoMarcos André GonçalvesThierson Couto RosaWellington Santos MartinsPublished in: ACM Trans. Inf. Syst. (2019)
Keyphrases
- learning to rank
- risk sensitive
- evolutionary multi objective
- feature selection
- optimization algorithm
- optimal control
- ranking functions
- utility function
- markov decision processes
- multi objective genetic algorithm
- loss function
- information retrieval
- support vector
- feature set
- model free
- text categorization
- optimality criterion
- ranking svm
- document retrieval
- text classification
- multi objective
- retrieval systems
- control policies
- mutual information
- machine learning
- information gain
- collaborative filtering
- decision theoretic
- markov decision problems
- expected utility
- support vector machine
- optimization method
- decision problems
- classification accuracy
- feature space
- differential evolution
- test collection
- dynamic programming
- training data
- particle swarm optimization