We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

stat.AP

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Statistics > Applications

Title: Anchor Attention for Hybrid Crowd Forecasts Aggregation

Abstract: Forecasting the future is a notoriously difficult task. To overcome this challenge, state-of-the-art forecasting platforms are "hybridized", they gather forecasts from a crowd of humans, as well as one or more machine models. However, an open challenge remains in how to optimally combine forecasts from these pools into a single forecast. We proposed anchor attention for this type of sequence summary problem. Each forecast is represented by a trainable embedding vector, and use computed anchor attention score as the combined weight. We evaluate our approach using data from real-world forecasting tournaments, and show that our method outperforms the current state-of-the-art aggregation approaches.
Subjects: Applications (stat.AP); Multiagent Systems (cs.MA)
Cite as: arXiv:2003.12447 [stat.AP]
  (or arXiv:2003.12447v1 [stat.AP] for this version)

Submission history

From: Yuzhong Huang [view email]
[v1] Tue, 3 Mar 2020 23:39:02 GMT (6435kb,D)

Link back to: arXiv, form interface, contact.