Timeline
11/10/17:
- 12:56 Changeset [15470] by
-
#2847 worked on M5Regression
- 09:05 Changeset [15469] by
-
#2789 trying to get SBART to work correctly.
11/09/17:
- 18:03 Changeset [15468] by
-
#2789 worked on sbart spline
- 16:23 Changeset [15467] by
-
#2789 dllexport bvalue function
- 11:51 Changeset [15466] by
-
#2809: Simplified the overall filtering logic as suggested by bburlacu
- changed parameter names to actively reflect that filter means "remaining"
- moved filter combination logic to FilterContent
- simplified/restructured code
11/08/17:
- 20:23 Changeset [15465] by
-
Started a summary of findings.
- 17:01 Changeset [15464] by
-
#2854: Corrected results and parameter assignment of random forest model for regression and classification.
- 17:00 Ticket #2854 (Random Forest Results are wrong) created by
- Out-of-bag error estimates are significantly smaller than standard …
- 16:32 Changeset [15463] by
-
#2817
-Added unit tests for bin packing 3d.
-The items can now be sorted by the material.
- 16:30 Changeset [15462] by
-
#2817
-Added unit tests for bin packing 3d.
-The items can now be sorted by the material.
- 12:16 Changeset [15461] by
-
#2853: Reduced test set size for Keijzer 11 - 15.
- 12:10 Ticket #2853 (Reduce test set size of Keijzer problem instances) created by
- Currently some Keijzer benchmark instances contain more than 300k …
- 07:34 Changeset [15460] by
-
#2747: worked on the CFSAP
- Added problem definition that defines both sequence and assignment for a single nest
- Added problem definition that would optimizes both sequence and assignment for multiple nests
- Added interface
- Added solving strategy that would use multiple instances of a template algorithm to optimize the worst nest
- Fixed bug in parser regarding setup times
11/07/17:
- 19:49 Changeset [15459] by
-
experimentation
- 17:23 Changeset [15458] by
-
#2789 recomiled sbart
- 17:03 Changeset [15457] by
-
#2789 added Finbarr O'Sullivan smoothing spline code
- 13:49 Changeset [15456] by
-
#2747: worked on the CFSAP
- Introduced new benchmark instances mentioned in the literature and updated the parser
- 13:15 Changeset [15455] by
-
#2847 added WeightedEuclideanDistance && fixed minor bug in scatterPlot colloring
- 08:24 Changeset [15454] by
-
- Extreme point bin packing does not need the occupation layer anymore
- Changes at the fitting algorithm. Now they are packing the items as in the paper of S. Martello, D. Pisinger, D. Vigo described
11/06/17:
- 18:20 AdditionalMaterial edited by
- added ijspm 2017 (diff)
- 18:19 Changeset [15453] by
-
Uploaded additional material for IJSPM (uncertainty analysis)
- 16:38 Changeset [15452] by
-
- Removed ManualResetEventSlim allOptimizerFinished.
- Reworked execution state transitions into simple, documented rules
- Fixed potentiall NullReferenceException in ExperimentView
- 15:57 Changeset [15451] by
-
#2850 created branch & added WeightedEuclideanDistance
- 14:18 Documentation/DevelopmentCenter/DeveloperGuidelines edited by
- (diff)
- 13:12 Changeset [15450] by
-
#2789 more tests with CV and automatic determination of smoothing parameter
11/05/17:
11/03/17:
- 20:28 Changeset [15449] by
-
#2789 created x64 build of CUBGCV algorithm, fixed some bugs
- 17:07 Changeset [15448] by
-
#2852: Added counts of function and gradient evaluations by constants optimization as results.
- 15:28 Changeset [15447] by
-
#2852: Adapted constants optimization and auto diff converter to not add linear scaling terms.
- 15:14 Ticket #2852 (Improve Constants Optimization for Symbolic Regression) created by
- * Constants optimization should work without linear scaling * …
- 11:14 Changeset [15446] by
-
- Start progres with initial progress value in several cases to use the progress value update.
- Removed object disposal guard in ProgressView because it can block part of the initialization.
- 09:45 Changeset [15445] by
-
#2845: Removed progressbar state (green/orange/red).
11/02/17:
- 16:55 Changeset [15444] by
-
#2845 Cancel the artificial delay if progress is stopped.
11/01/17:
- 18:05 Changeset [15443] by
-
#2789 added interface to CUBGCV using DllImport of Fortran library
10/31/17:
- 20:01 Changeset [15442] by
-
#2789: added cross-validation and automatic determination of regularization parameter (naive CV is slow!).
Research produced ACM Algorithm 642.f (Fortran) for univariate cubic spline smoothing with generalized cross-validation in O(n) described in HUTCHINSON, M. F., AND DE HOOG, F.R. Smoothing noisy data with spline functions. Numer. Math. 4 7 (1985), 99-106. (TODO: translate to C#)
10/29/17:
- 22:22 Changeset [15441] by
-
#2796 more bug fixing
- 21:22 Changeset [15440] by
-
#2796 fixed hashing of expressions and unit tests for expression enumeration
10/28/17:
- 20:32 Changeset [15439] by
-
#2796 more refactoring and working on unit tests
- 19:56 Changeset [15438] by
-
#2796 refactoring to simplify the code
