All Versions
23
Latest Version
Avg Release Cycle
61 days
Latest Release
1231 days ago
Changelog History
Page 2
Changelog History
Page 2
-
v2.0.0 Changes
November 22, 2019Smile has been fully rewritten with more than 150,000 lines change.
- Fully redesigned API. It is leaner, simpler and even more friendly.
- Faster implementation and memory optimization. Many algorithms are fully reimplemented. RandomForest is 8X faster than XGBoost on large benchmark data (10MM samples).
- ๐ New parallelism mechanism
- All new DataFrame and Formula
- ๐ New algorithms such as ICA, error reduction prune, quantile loss, TWCNB, etc.
- ๐ Support arbitrary class labels.
- โจ Enhancement and harden numeric computations.
- ๐ Support Parquet, SAS, Arrow, Avro, etc.
- ๐ Bug fixes.
-
v2.0.0-RC2 Changes
November 13, 2019Smile has been fully rewritten with more than 100,000 lines change.
- Fully redesigned API. It is leaner, simpler and even more friendly.
- Faster implementation and memory optimization. Many algorithms are fully reimplemented. RandomForest is 8X faster than XGBoost on large benchmark data (10MM samples).
- ๐ New parallelism mechanism
- All new DataFrame and Formula
- ๐ New algorithms such as ICA, error reduction prune, quantile loss, TWCNB, etc.
- ๐ Support arbitrary class labels.
- โจ Enhancement and harden numeric computations.
- ๐ Support Parquet, SAS, Arrow, Avro, etc.
- ๐ Bug fixes.
-
v2.0.0-RC1 Changes
October 29, 2019- All new DataFrame and Formula
- ๐ Support Parquet, SAS, Arrow, Avro, etc.
- ๐ Performance and memory optimization
Smile has been fully rewritten with more than 100,000 lines change. Therefore, the API is not compatible with v1.
-
v1.5.3 Changes
June 02, 2019- ElasticNet
- GroupKFold
- ๐ Bug fixes
-
v1.5.2 Changes
October 15, 2018- K-Modes clustering
- Online learning with LogisticRegression by SGD
- MCC (Matthews correlation coefficient) metric
- ๐ Bug fixes
-
v1.5.1 Changes
February 26, 2018๐ 1. Performance improvement of hierarchical clustering ๐ 2. Bug fixes.
-
v1.5.0 Changes
November 10, 2017- DataFrame ๐ง 2. New Shell for Mac and Linux ๐ 3. Shell improvement for Windows ๐ 4. Out of box support of native LAPACK for Windows
- Scala functions to export AttributeDataset, double[][], double[] to ARFF or CSV
- Scala functions for validation measures ๐จ 7. Refactor feature transformation and generation classes
- NeuralNetwork for regression
- Recursive least squares ๐จ 10. Refactor Scala NLP API ๐ 11. Bug fixes
-
v1.5.0-RC3
November 06, 2017 -
v1.5.0-RC2
October 29, 2017 -
v1.5.0-RC1
September 28, 2017