Share This Author
Estimation-Action-Reflection: Towards Deep Interaction Between Conversational and Recommender Systems
This work proposes a new CRS framework named Estimation"Action" Reflection, or EAR, which consists of three stages to better converse with users, and conducts extensive experiments on two datasets from Yelp and LastFM to demonstrate significant improvements over the state-of-the-art method CRM.
Learning Hidden Features for Contextual Bandits
This paper rigorously proves that the developed contextual bandit algorithm achieves a sublinear upper regret bound with high probability, and a linear regret is inevitable if one fails to model such hidden features.
Contextual Bandits in a Collaborative Environment
This paper develops a collaborative contextual bandit algorithm, in which the adjacency graph among users is leveraged to share context and payoffs among neighboring users while online updating, and rigorously proves an improved upper regret bound.
Factorization Bandits for Interactive Recommendation
A high probability sublinear upper regret bound is proved for the developed algorithm, where considerable regret reduction is achieved on both user and item sides.
Learning Contextual Bandits in a Non-stationary Environment
- Qingyun Wu, N. Iyer, Hongning Wang
- Computer ScienceAnnual International ACM SIGIR Conference on…
- 23 May 2018
This paper proposes a contextual bandit algorithm that detects possible changes of environment based on its reward estimation confidence and updates its arm selection strategy respectively and demonstrates its learning effectiveness in such a non-trivial environment.
FLAML: A Fast and Lightweight AutoML Library
- Chi Wang, Qingyun Wu, Markus Weimer, Erkang Zhu
- Computer ScienceConference on Machine Learning and Systems
- 12 November 2019
A fast and lightweight library FLAML is built which optimizes for low computational resource in finding accurate models and significantly outperforms top-ranked AutoML libraries on a large open source AutoML benchmark under equal, or sometimes orders of magnitude smaller budget constraints.
Seamlessly Unifying Attributes and Items: Conversational Recommendation for Cold-start Users
- Shijun Li, Wenqiang Lei, Qingyun Wu, Xiangnan He, Peng Jiang, Tat-Seng Chua
- Computer ScienceACM Trans. Inf. Syst.
- 23 May 2020
The Conversational Thompson Sampling (ConTS) model holistically solves all questions in conversational recommendation by choosing the arm with the maximal reward to play and seamlessly unify attributes and items in the same arm space and achieve their EE trade-offs automatically using the framework ofThompson Sampling.
Factorization Bandits for Online Influence Maximization
- Qingyun Wu, Zhige Li, Huazheng Wang, Wei Chen, Hongning Wang
- Computer ScienceKnowledge Discovery and Data Mining
- 9 June 2019
This paper proposes an upper confidence bound based online learning solution to estimate the latent factors, and therefore the activation probabilities, of online influence maximization in social networks, and factorizes the activation probability on the edges into latent factors on the corresponding nodes.
Dynamic Ensemble of Contextual Bandits to Satisfy Users' Changing Interests
This work capitalize on the unique context-dependent property of reward changes to conquer the challenging non-stationary environment for model update and provides a rigorous upper regret bound analysis of the proposed algorithm.
Variance Reduction in Gradient Exploration for Online Learning to Rank
- Huazheng Wang, Sonwoo Kim, Eric McCord-Snook, Qingyun Wu, Hongning Wang
- Computer ScienceSIGIR
- 10 June 2019
This work projects the selected updating direction into a space spanned by the feature vectors from examined documents under the current query, after an interleaved test, and proves that this projected gradient is still an unbiased estimation of the true gradient, and shows that this lower-variance gradient estimation results in significant regret reduction.