A Large-Scale Neutral Comparison Study of Survival Models on Low-Dimensional Data
This site aggregates additional results and metadata accompanying the benchmark project.
All code is available on GitHub.
Note that this page belongs to the most recent version of the project.
A previous version from 2024 is available here.
Downloading Results
Results can be downloaded from the web.
Files are suffixed with the associated tuning measure unless already combined, where harrell_c,isbs indicates results for learners which are not at all or not explicitly tuned during the experiment but rather use an internal mechanism, i.e. Kaplan-Meier or CoxBoost (see learner table on Metadata)
scoresare outer evaluation scores, i.e., one score pertaskxlearnerxtuning_measurexiter(outer resampling iteration)aggrare aggregates scores, i.e., one score pertaskandlearnerandtuning_measurebma_(BenchmarkAggr) are intermediate results and are only relevant if thesemlr3objects are specifically required.
See also the table of measures for a list of measures and their abbreviated and full names.
| File | Description |
|---|---|
scores.rds |
Scores for all learners, tasks and measures for individual outer resampling iterations |
aggr.rds |
Aggregated scores (averaged across outer resampling folds) per task and learner |
archives.rds |
All tuning archives produced during the benchmark as an RDS file |
archives.zip |
ZIP file of all tuning archives (over 19000) in CSV files for each combination of learner, task, tuning measure, and resampling iteration |