P
Padhraic Smyth
Researcher at University of California, Irvine
Publications - 359
Citations - 38795
Padhraic Smyth is an academic researcher from University of California, Irvine. The author has contributed to research in topics: Inference & Topic model. The author has an hindex of 80, co-authored 342 publications receiving 36653 citations. Previous affiliations of Padhraic Smyth include University of California & Jet Propulsion Laboratory.
Papers
More filters
Posted Content
Variational Beam Search for Online Learning with Distribution Shifts.
TL;DR: A new Bayesian meta-algorithm is proposed that can both make inferences about subtle distribution shifts based on minimal sequential observations and accordingly adapt a model in an online fashion and yields significant improvements over state-of-the-art Bayesian online learning approaches.
Zero-Shot Batch-Level Anomaly Detection
TL;DR: Adaptive Centered Representation (ACR) as mentioned in this paper is a simple yet effective method for zero-shot batch-level anomaly detection, which trains off-the-shelf deep anomaly detectors to adapt to a set of inter-related training data distributions in combination with batch normalization.
Feature Model for Longitudinal Social Networks
TL;DR: It is shown how the number of features and their trajectories for each actor can be inferred simultaneously and the utility of this model on prediction tasks using synthetic and real-world data is demonstrated.
Journal ArticleDOI
Climate-driven changes in the predictability of seasonal precipitation
Phong V. V. Le,James T. Randerson,Stephen J. Wright,Padhraic Smyth,Clement Guilloteau,Antonios Mamalakis,Efi Foufoula-Georgiou +6 more
TL;DR: In this article , on the basis of CMIP6 models that capture the present-day teleconnections between seasonal precipitation and previous-season sea surface temperature (SST), the authors show that climate change is expected to alter the SST-precipitation relationships and thus their ability to predict seasonal precipitation by 2100.
Proceedings Article
Bayesian Predictive Profiles With Applications to Retail Transaction Data
Igor V. Cadez,Padhraic Smyth +1 more
TL;DR: A generative mixture model for count data is described and an an approximate Bayesian estimation framework that effectively combines an individual's specific history with more general population patterns is used.