@Dmytro:
- Ready mag page MUST be accessible at all times with table of current performances. Each application must include current AUC, baseline AUC, n-total, Prediction window, inference window, precursor phenotypes, target phenotypes
- Formalize the parameters of the general model. The objective here is to describe the xCOR model as a tuple: $$ M(I,P,C,\rho_h,\tau_h): \mathcal{D} \rightarrow [0,1] $$ where \( I,P,C\) are the inference, prediction and confidence windows, \( \mathcal{D}\) is the dataset, and the mapping is the risk calculation. Lets identify if our applications conform to thsi model, extend it if necessary (in an elegant manner), and maintain a table with all the parameters. This is the basis of the general xCor paper
- Complete pulmonary fibrosis and manic 2 publisher files
@Yi
- Please finish the DS2 update, with the perturbation and approximate metrization result
- For the file Influenza Variation Data use ds2 to find simialrities between counties. Find clusters of simialr counties and produce a geospatial map for it.
- Try also Staph infection, Strep infection and Ecoli infection.
@Victor
- Report of Cardiac analysis [vary test/train split to get over baseline]
- Start terrorism analysis gcd dataset
- Do perturbation anlysis for N1, N3, N7 and augemnet perturbation matrices for all cities
@Lyn
- Please share definitive results on mubiome analysis. Need to see a google doc with results please
@david
- report of qnet construction on longitudinal survey
- PLEASE LEARN SLURM
- Schedule midweek meeting
@we
- will discuss the pandemic 2 paper with @jin
- Will share draft for midwek meeting
@jin
- creed analysis to be discussed
- New task for Time Aware Machine Learning (TAMI). See cartoon
@alice
- will discuss ML docx