R
Ran El-Yaniv
Researcher at Technion – Israel Institute of Technology
Publications - 138
Citations - 14744
Ran El-Yaniv is an academic researcher from Technion – Israel Institute of Technology. The author has contributed to research in topics: Support vector machine & Competitive analysis. The author has an hindex of 40, co-authored 133 publications receiving 12684 citations. Previous affiliations of Ran El-Yaniv include University of Toronto & Google.
Papers
More filters
Proceedings ArticleDOI
Multi-Hop Paragraph Retrieval for Open-Domain Question Answering
Yair Feldman,Ran El-Yaniv +1 more
TL;DR: A method for retrieving multiple supporting paragraphs, nested amidst a large knowledge base, which contain the necessary evidence to answer a given question, by forming a joint vector representation of both a question and a paragraph.
Journal ArticleDOI
Season of birth, natural light, and myopia.
Yossi Mandel,Itamar Grotto,Itamar Grotto,Ran El-Yaniv,Michael Belkin,Eran Israeli,Eran Israeli,Uri Polat,Elisha Bartov +8 more
TL;DR: Myopia in this population is associated with birth during summer months and the exact associating mechanism is not known but might be related to exposure to natural light during the early perinatal period.
Book ChapterDOI
Transductive Rademacher complexity and its applications
Ran El-Yaniv,Dmitry Pechyony +1 more
TL;DR: A new PAC-Bayesian bound for mixtures of transductive algorithms based on a new Rademacher bound based on the "unlabeled-labeled" decomposition technique applies to many current and practical graph-based algorithms.
Journal ArticleDOI
Transductive Rademacher Complexity and its Applications
Ran El-Yaniv,Dmitry Pechyony +1 more
TL;DR: A technique for deriving data-dependent error bounds for transductive learning algorithms based on transductIVE Rademacher complexity is developed and a new PAC-Bayesian bound is presented for mixtures of transductives based on a novel general error bound for transduction.
Book ChapterDOI
Iterative Double Clustering for Unsupervised and Semi-Supervised Learning
Ran El-Yaniv,Oren Souroujon +1 more
TL;DR: It is demonstrated that the IDC algorithm is especially advantageous when the data exhibits high attribute noise, and a natural extension of IDC for (semi-supervised) transductive learning where it is given both labeled and unlabeled examples is proposed.