Open vinaypedapati opened 9 months ago
See if ELN classification gives ~% chance of survival, then test that against the data. See if our model has a prediction that is closer to the actual survival % at ... years.
Need to double check the literature.
Testing the AML paper's model on the data and comparing its results to the ones we are getting.
Testing PPV, NPV and other error metrics
Determining whether our results are "significantly" different from ELN's model
Spent the day watching videos on stat quest about ADA boost, ridge and Lasso regression, gradient boosting, and xgboost algorithm
Will try to implement xgboost by next week