Gradual Learning and Convergence

By Joe Pater. Linguistic Inquiry, 2008

109 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Cultural Studies

View Paper: https://www.semanticscholar.org/paper/2028ce655197a839393f7c3edc50ba6fb9c3b198

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Kie Zuraw
2. Bruce Hayes
3. Joe Pater
4. KevinM . Ryan
5. Andries W. Coetzee
6. D. C. Hall

In Cultural Studies

1. Bruce Hayes
2. Kie Zuraw

Citations in the corpus

• Intersecting constraint families: An argument for harmonic grammar (2017, methodology): “The gradual learning algorithm (GLA; Boersma 1998a) can be used to learn stochastic OT grammars, although as Pater (2008) discusses, the algorithm is not guaranteed to converge, and even if it does converge the result is not guaranteed to be the stochastic OT grammar that best matches the data.”
• Weighted Constraints in Generative Linguistics (2009, methodology): “…on the size of windows can be obtained by imposing a maximum value on the weights and a minimum difference between the harmony scores of the optima and their competitors (see Boersma & Pater, 2008 and Potts et al., 2009 on the margin of separation of harmony; see relatedly Albright et al., 2008).”