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


Current browse context:


Change to browse by:


References & Citations

DBLP - CS Bibliography


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

Computer Science > Computation and Language

Title: On the importance of normative data in speech-based assessment

Abstract: Data sets for identifying Alzheimer's disease (AD) are often relatively sparse, which limits their ability to train generalizable models. Here, we augment such a data set, DementiaBank, with each of two normative data sets, the Wisconsin Longitudinal Study and Talk2Me, each of which employs a speech-based picture-description assessment. Through minority class oversampling with ADASYN, we outperform state-of-the-art results in binary classification of people with and without AD in DementiaBank. This work highlights the effectiveness of combining sparse and difficult-to-acquire patient data with relatively large and easily accessible normative datasets.
Subjects: Computation and Language (cs.CL)
Cite as: arXiv:1712.00069 [cs.CL]
  (or arXiv:1712.00069v1 [cs.CL] for this version)

Submission history

From: Chloé Pou-Prom [view email]
[v1] Thu, 30 Nov 2017 20:36:50 GMT (15kb)

Link back to: arXiv, form interface, contact.