/thesis

K08 & PhD

Primary LanguageJupyter Notebook

            1. Be/RLHF
                      \
   2. Do/Langugage -> 4. Be/Compute -> 5. Do/Training -> 6. Hide/Environment
                      /
                       3. Hide/Agent
  1. Personalized medicine
    • Literature has thus far been limited (as if older adults are homogenious)
    • Patient or Case-mix dependent inference (individual consent derived from single parameter)
    • Sort of Public Health style (clinical trial literature is no different)
  2. WebApp
    • Demographics (age, etc)
    • Exam (frailty phenotype)
    • Labs (inflammatory markers)
  3. Deployment
    • Not a onetime thing (i.e., not truth, frequentist)
    • Iterations (i.e., freq updates, bayesian)
    • Neural networks capable of RLHF (credibility intervals)
  4. RDC/NCHS Access Tokens
    • Compute: e(b), e(V) sizes
    • Iterations
    • Updates
  5. Vince
  6. GitHub
    • Transparency
    • Free
    • .js & .html vs. strimlit