Estimating the Hessian Matrix of Ranking Objectives for Stochastic Learning to Rank with Gradient Boosted Trees.
Jingwei KangMaarten de RijkeHarrie OosterhuisPublished in: CoRR (2024)
Keyphrases
- learning to rank
- hessian matrix
- ranking functions
- information retrieval
- ranking svm
- step size
- learning to rank algorithms
- evaluation measures
- loss function
- document retrieval
- query dependent
- direct optimization
- collaborative filtering
- ranking models
- gaussian derivatives
- decision trees
- ranking algorithm
- ranking list
- retrieval systems
- test collection
- image classification
- information retrieval systems
- web search
- supervised learning
- cost function
- optical flow
- image processing