Houjun Liu

Sadeghian 2021

# ntj

DOI: 10.3389/fcomp.2021.624594

(Sadeghian, Schaffer, and Zahorian 2021)


Using a genetic algorithm, picked features to optimize fore; achieved \(94\%\) with just MMSE data alone (ok like duh me too). Developed ASR tool to aid.


  • Developed an ASR methodology for speech, complete with punctuations
  • Used a genetic algorithm to do feature selection; NNs performed worse because “space is smaller???”

Notable Methods

Used a GRU to insert punctuations

The paper leveraged the nuke that is a bidirectional GRU, ATTENTION,

Key Figs

Fully automated ANN transcript does pretty well in terms of classifier AD/NL.

New Concepts


very confusing (too many things going on at once)