English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

We study and quantify the generalization patterns of multitask learning (MTL)models for sequence labeling tasks. MTL models are trained to optimize a set ofrelated tasks jointly. Although multitask learning has achieved improvedperformance in some problems, there are also tasks that lose performance whentrained together. These mixed results motivate us to study the factors thatimpact the performance of MTL models. We note that theoretical bounds andconvergence rates for MTL models exist, but they rely on strong assumptionssuch as task relatedness and the use of balanced datasets. To remedy theselimitations, we propose the creation of a task simulator and the use ofSymbolic Regression to learn expressions relating model performance to possiblefactors of influence. For MTL, we study the model performance against thenumber of tasks (T), the number of samples per task (n) and the taskrelatedness measured by the adjusted mutual information (AMI). In ourexperiments, we could empirically find formulas relating model performance withfactors of sqrt(n), sqrt(T), which are equivalent to sound mathematical proofsin Maurer[2016], and we went beyond by discovering that performance relates toa factor of sqrt(AMI).

Learning Functions to Study the Benefit of Multitask Learning