Releases: SmartDataAnalytics/SML-Bench
Releases · SmartDataAnalytics/SML-Bench
SML-Bench 0.3.0
- Learning scenarios
- Updates in mammographic dataset
- Updates and bugfix in nctrer dataset
- Bugfix in suramin dataset
- Learning systems
- Added SWI Prolog port of Aleph
- Minor updates in Aleph scripts
SML-Bench 0.2
- Framework
- Support for direct learning system configurations inside the benchmark configuration via e.g.
learningsystems.aleph.construct_bottom = saturation
- Support to run multiple instances of a learning system with different configurations
- Support for nested learning system configurations inside the the benchmark configuration
- Maven profile for code quality feedback
- Switched to Java 8
- Core
- Refactoring
- Increased test coverage
- Increased code documentation
- Cleaned up configuration handling
- Stability improvements and bug fixes
- MEX support
- Refactoring
- Bug fixes
- Benchmarking result output
- Added CSV output writer
- Support for direct learning system configurations inside the benchmark configuration via e.g.
- Learning scenarios
- Extended dataset information in datasets.ttl file
- Added animals scenarios
- Bugfix in pyrimidine dataset
- Bugfix in lymphography dataset
- Learning systems
- Experimental support for TreeLiker
- Updates in the learning system wrappers to support direct configurations inside the benchmark configuration
- Updated/fixed language bias declarations
- Bugfixes in learning system wrappers
SML-Bench 0.1
- Basic framework functionality
- N-fold cross validation
- Benchmark report
- Detailed logging of the tool settings and outputs
- Initial MEX support
- Supported learning systems
- Aleph
- DL-Learner
- GILPS comprising
- FuncLog
- ProGolem
- TopLog
- Golem
- Progol
- Learning problems for the following datasets provided
- Carcinogenesis
- Hepatitis
- Lymphography
- Mammographic
- Mutagenesis
- NCTRER
- Premierleague
- Pyrimidine
- Suramin