SutraNets: Sub-series Autoregressive Networks for Long-Sequence, Probabilistic Forecasting

no code implementations • NeurIPS 2023 • Shane Bergsma, Timothy Zeyl, Lei Guo

We find SutraNets to significantly improve forecasting accuracy over competitive alternatives on six real-world datasets, including when we vary the number of sub-series and scale up the depth and width of the underlying sequence models.

Time Series

C2FAR: Coarse-to-Fine Autoregressive Networks for Precise Probabilistic Forecasting

1 code implementation • 22 Dec 2023 • Shane Bergsma, Timothy Zeyl, Javad Rahimipour Anaraki, Lei Guo

We present coarse-to-fine autoregressive networks (C2FAR), a method for modeling the probability distribution of univariate, numeric random variables.

Anomaly Detection Time Series

I'm a Belieber: Social Roles via Self-identification and Conceptual Attributes

no code implementations • ACL 2014 • Charley Beller, Rebecca Knowles, Craig Harman, Shane Bergsma, Margaret Mitchell, Benjamin Van Durme

Using Conceptual Class Attributes to Characterize Social Media Users

no code implementations • ACL 2013 • Shane Bergsma, Benjamin Van Durme

Common Sense Reasoning

Explicit and Implicit Syntactic Features for Text Classification

no code implementations • ACL 2013 • Matt Post, Shane Bergsma

General Classification Language Modelling +2

Broadly Improving User Classification via Communication-Based Name and Location Clustering on Twitter

no code implementations • NAACL 2013 • Shane Bergsma, Mark Dredze, Benjamin Van Durme, Theresa Wilson, David Yarowsky

Clustering General Classification

Stylometric Analysis of Scientific Articles

no code implementations • NAACL 2012 • Shane Bergsma, Matt Post, David Yarowsky

Language Identification for Creating Language-Specific Twitter Collections

no code implementations • WS 2012 • Shane Bergsma, Paul McNamee, Mossaab Bagdouri, Clayton Fink, Theresa Wilson

Language Identification