Abstract | ||
---|---|---|
We consider the task of assigning experts from a port- folio of specialists in order to solve a set of tasks. We apply a Bayesian model which combines collaborative filtering with a feature-based description of tasks and experts to yield a general framework for managing a portfolio of experts. The model learns an embedding of tasks and problems into a latent space in which affin- ity is measured by the inner product. The model can be trained incrementally and can track non-stationary data, tracking potentially changing expert and task character- istics. The approach allows us to use a principled de- cision theoretic framework for expert selection, allow- ing the user to choose a utility function that best suits their objectives. The model component for taking into account the performance feedback data is pluggable, al- lowing flexibility. We apply the model to manage a port- folio of algorithms to solve hard combinatorial prob- lems. This is a well studied area and we demonstrate a large improvement on the state of the art in one do- main (constraint solving) and in a second domain (com- binatorial auctions) created a portfolio that performed significantly better than any single algorithm. |
Year | Venue | Keywords |
---|---|---|
2010 | AAAI | bayesian model,machine learning,linear programming,inner product,collaborative filtering,portfolio management,recommender system |
Field | DocType | Citations |
Collaborative filtering,Embedding,Bayesian inference,Project portfolio management,Computer science,Combinatorial auction,Portfolio,Artificial intelligence,Performance feedback,Machine learning | Conference | 13 |
PageRank | References | Authors |
0.73 | 15 | 6 |
Name | Order | Citations | PageRank |
---|---|---|---|
David H. Stern | 1 | 143 | 9.99 |
Horst Samulowitz | 2 | 316 | 26.05 |
Ralf Herbrich | 3 | 1660 | 170.61 |
Thore Graepel | 4 | 4211 | 242.71 |
Luca Pulina | 5 | 326 | 37.95 |
Armando Tacchella | 6 | 1448 | 108.82 |