Skip to content

Latest commit

 

History

History
23 lines (14 loc) · 735 Bytes

README.md

File metadata and controls

23 lines (14 loc) · 735 Bytes

AdaRank

A python implementation of the AdaRank algorithm. It is based on weak ranker which use signle feature.

How to Run

The following code will run Adarank for 500 iteration with optimzation function NDCG@5. If all features are selected 5 times, our algorithm will stop.

model = Adaank(scorer=NDCGScorer_qid(K=5))
model.fit(X, y, qid, X_vali, y_vali, qid_vali)
pred = model.predict(X_test)
print(NDCGScorer_qid(K=5)(y_test,pred,qid_test).mean())

References

Xu and Li. AdaRank: a boosting algorithm for information retrieval. In Proceedings of SIGIR '07, pages 391–398. ACM, 2007.

rueycheng/AdaRank

The Lemur Project