Modeling Similarities Among Autism Spectrum Patients Using Word Embeddings on Clinical Notes

<p>Autism Spectrum Disorder (ASD) is characterized by difficulties in areas of social communication, reciprocal social interaction, restricted or repetitive patterns of behavior and interests, and cognitive or significant delays in early language development. Although we are seeing consistent research being done on understanding the genetic and biological aspects of ASD, diagnosing ASD patients is solely based on behavioral symptoms.</p> <p>In this thesis, we leverage unsupervised machine learning techniques to better understand ASD patients and the challenging behaviors they present. We used Doc2Vec to create neural word embedding vectors on the clinical notes presented and K-means clustering to group the patients based on similarities in the notes. The clusters will give us greater insight into the examinations done by clinicians in ABA therapy, the challenging behaviors presented, and the similarities between patients in the cluster.</p>

Name thesis:degree_name

Master of Science (MS)

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Computational and Data Sciences

Year dc:date.available

2022

Author dc:creator

• Pirzadeh, Raha

Contributors dc:contributor

• Erik Linstead
• Elia Eiroa Lledo
• Dennis Dixon

Repository record dc:identifier

https://digitalcommons.chapman.edu/cads_theses/15

OAI identifier oai:identifier

oai:digitalcommons.chapman.edu:cads_theses-1014