Y
Yuexin Wu
Researcher at Carnegie Mellon University
Publications - 40
Citations - 2017
Yuexin Wu is an academic researcher from Carnegie Mellon University. The author has contributed to research in topics: Computer science & Search algorithm. The author has an hindex of 15, co-authored 27 publications receiving 1421 citations. Previous affiliations of Yuexin Wu include Microsoft & Tsinghua University.
Papers
More filters
Posted ContentDOI
A Deep Boosting Based Approach for Capturing the Sequence Binding Preferences of RNA-Binding Proteins from High-Throughput CLIP-Seq Data
TL;DR: A deep boosting based machine learning approach, called DeBooster, to accurately model the binding sequence preferences and identify the corresponding binding targets of RBPs from CLIP-seq data, and may provide an effective index to investigate the effect of pathogenic mutations in RBP binding sites.
Journal ArticleDOI
Conditional Adapters: Parameter-efficient Transfer Learning with Fast Inference
Tao Lei,Junwen Bai,Siddhartha Brahma,Joshua Ainslie,Ken Wing Kuen Lee,Yaoqi Zhou,Nan Du,Vincent Zhao,Yuexin Wu,Bo Li,Yu H. Zhang,Ming-Wei Chang +11 more
TL;DR: Conditional Adapter (CoDA) as discussed by the authors generalizes beyond standard adapter approaches to enable a new way of balancing speed and accuracy using conditional computation and achieves a 2x to 8x inference speedup compared to the state-of-the-art Adapter approach with moderate to no accuracy loss and the same parameter efficiency.
Book ChapterDOI
Computational Protein Design Using AND/OR Branch-and-Bound Search
TL;DR: This paper proposes a new protein design algorithm based on the AND/OR branch-and-bound (AOBB) search, which is a variant of the traditional branch- and-bound search algorithm, to solve this combinatorial optimization problem.
Posted Content
Computational Protein Design Using AND/OR Branch-and-Bound Search
TL;DR: A branch-and-bound (AOBB) search algorithm was proposed in this paper to solve the problem of global minimum energy conformation (GMEC) in protein design, which can provide a large speedup by several orders of magnitude while still guaranteeing to find the solution.
Proceedings ArticleDOI
Contextual Encoding for Translation Quality Estimation.
TL;DR: This paper proposed a method to effectively encode the local and global contextual information for each target word using a three-part neural network approach, which achieves strong results, ranking first in three of the six tracks.