FastProp Benchmarks
A compilation of benchmark notebooks that demonstrate and compare the performance of getML's FastProp algorithm against various implementations of other propositionalization algorithms across different datasets.
Note
FastProp (short for fast propositionalization) is an open source Feature Learner and available in the getML Community Edition.
Results
| Faster vs. featuretools | Faster vs. tsfresh | Remarks | |
|---|---|---|---|
| Air pollution | ~65x | ~33x | The predictive accuracy can be significantly improved by using RelMT instead of propositionalization approaches, please refer to this notebook. |
| Dodgers | ~42x | ~75x | The predictive accuracy can be significantly improved by using the mapping preprocessor and/or more advanced feature learning algorithms, please refer to this notebook. |
| Interstate94 | ~55x | ||
| Occupancy | ~87x | ~41x | |
| Robot | ~162x | ~77x |
Source
These notebooks are published on the getml-demo repository.