This is an old revision of the document!
Introduction
Ranking is an important concept in modelling our preferences.
We rank hotels by their quality using one star to five stars;
we rank baseball teams by their records using pairwise competitions;
we rank job applicants by their ability using ordered scores.
In machine learning, the ranking concept
corresponds to a rich family of important problems,
which lend themselves to
a wide range of applications from social science to
behavioural science to information retrieval.
For instance, in a Web search system,
we want the machines to automatically
rank/order the results of our query based on relevance;
in an online shopping system,
we want the machines to automatically
rank/rate the products based on user evaluations;
in a music playing system,
we want the machines to automatically
rank/recommend the songs based on our personal tastes.
paper
Yu-Xun Ruan, Hsuan-Tien Lin and Ming-Feng Tsai, Improving Ranking Performance with Cost-sensitive Ordinal Classification via Regression, Improving Ranking Performance with Cost-sensitive Ordinal Classification via Regression. Accepted by Information Retrieval, 2013.
link
Hsuan-Tien Lin and Ling Li. Reduction from Cost-sensitive Ordinal Ranking to Weighted Binary Classification. Neural Computation, to appear. Some preliminary parts appeared in NIPS '06 and PL Workshop @ ECML/PKDD '09.
link
Ming-Feng Tsai, Shang-Tse Chen, Yao-Nan Chen, Chun-Sung Ferng, Chia-Hsuan Wang, Tzay-Yeu Wen and Hsuan-Tien Lin. An Ensemble Ranking Solution to the Yahoo! Learning to Rank Challenge. National Taiwan University, Technical Report, September 2010.
link
Hsuan-Tien Lin and Ling Li. Combining Ordinal Preferences by Boosting. Preference Learning Workshop @ ECML/PKDD '09, 2009.
link
thesis
Yu-Xun Ruan. Studies on Ordinal Ranking with Regression. Master's Thesis, 2010.
Ken-Yi Lin. Data Selection Techniques for Large-scale RankSVM. Master's Thesis, 2009.