: NLP Vocabulary: linear model

NLP methods and tools > model > statistical model > linear model

NLP methods and tools > statistics > statistical model > linear model

linear model

A model that assigns one weight per feature to make predictions. (Google for Developers, Machine Learning Glossary)

Baxter (2000) shows in the context of linear models that if two problems P and R share optimal hypothesis classes then the induction of a model from a sample of P can efficiently regularize the induction of a model from a sample of R. (S\\ogaard, 2017)
Having irrelevant features can decrease the accuracy of machine learning models especially when dealing with linear models such as support vector machines. (Beccaria, Gagliardi & Kokkinakis, 2022)
Hessel and Schofield (2021) also reported that BoW-BERT performed better than other BoW models on GLUE including linear models with TF-IDF features. (Mamakas, Tsotsi, Androutsopoulos & Chalkidis, 2022)
In addition we evaluated this linear model by training new parsers with one of the systems under investigation here on treebanks in the most recent release of UD that did not already have a model and obtained predictions that were not outlandish especially for higher performing treebanks. (Anderson & Gómez-Rodríguez, 2022)

http://data.loterre.fr/ark:/67375/8LP-N9HM6H5F-X

RDF/XML TURTLE JSON-LD Last modified 5/15/24

Loterre