Week 39

05/15/23 - 05/19/23

Done and To-Do

  • Admin work: rec letters for UROP + letter to Lisa Yang
  • Summary of MIT-CHOP medical meeting here - accessible only to Satra
  • High-level summary of BHS data in anticipation of Monday’s CHOP meeting
    • Demographics of population we’re working with (those with longitudinal data)
    • Dealing with imbalanced data: race/ethnicity, sex, age
    • What are our exclusion criteria?
    • How are we splitting up train and test data
    • Walk through technical steps (how are we building models, which features etc.)
    • What are the target variables we want to predict?
  • Finish UMAP and PCA on HBN responses
    • Look specifically at PC loadings across ADHD subtypes - can we identify cognitive profiles?
    • Do ADHD subtypes cluster into neat groups? not really…
  • ADHD Paper: methods + results
    • Model ABCD data (same questionnaires)