Timeline
08/21/12:
- 16:20 Changeset [8515] by
-
#1886 fixed cloning and resetting of the scatter plot helper
- 15:53 Changeset [8514] by
-
#1894: experimented with different settings in cost and heuristic function
- 14:10 Changeset [8513] by
-
#1758: merged r8023:8509 from trunk to branch to make it useable again
- 13:40 Changeset [8512] by
-
#1894: tweaking of max edge speeds and heuristic cost function
- 13:32 Changeset [8511] by
-
#1886 added an option so that the scatter plot helper can now also produce data tables
- 13:27 Changeset [8510] by
-
#1776: added AUC to ClassificationEnsembleSolutionAccuracyToCoveredSamples
- 11:05 Ticket #1920 (Models of an Ensemble may get a different problem data if loaded) created by
- The AfterDeserialization method in *EnsembleSolution give the …
- 10:23 Ticket #1919 (Classification ensemble cannot be loaded if no samples are in training) created by
- If an classification ensemble is created and the problem data is …
08/20/12:
- 17:32 Changeset [8509] by
-
#1894:
Dijkstra: get node with a specific rank
graph interface extended: get vertices
fixed bug in ReadData method
methods to perform routing algorithm benchmark added to test program
- 17:24 Changeset [8508] by
-
#1776: merged r7864:8507 trunk into branch
- 16:34 Changeset [8507] by
-
#1776: adapted ClassificationEnsembleSolutionAccuracyToCoveredSamples
- 13:00 Changeset [8506] by
-
- Web project now creates custom html for each type we want to enter / display
- Added endpointConfigurationName to HiveServiceLocator (because Web Project contains more than one endpoint configuration)
- Removed logging statement from ConfigurationService to prevent exception during failure of loading ConfigurationSettings
- ApplicationManager: Changed default implementation to WebApplicationManager (instead of LightWeight) for testing purposes within Web Project
- WebApplicationManager: The application manager which returns plugin descriptors from the currently loaded assemblies (instead of LightweightAppManager)
- HiveService: Fixed a transaction bug
- IControllerService: Created a method to dispatch Scenarios to certain IScenarioManager (in this case HiveScenarioManager)
- Added more mappable types to ControllerModel
- PlaceholderControllerService dispatches all Scenarios to the HiveScenarioManager
- Web project now dispatches the scenario to the controller after pressing "Run Job"
- 12:40 Changeset [8505] by
-
#1886 fixed cloning and storing
- 11:39 Changeset [8504] by
-
#1894: new read data method using NameTable for better performance
- 10:17 Changeset [8503] by
-
#1886 switched rows and columns of statistical tabular view
08/19/12:
- 20:34 Changeset [8502] by
-
#1886 added a run collection view that shows statistical information for data table rows
- 15:56 Changeset [8501] by
-
#1886 fixed cloning and made helper classes items
08/18/12:
- 21:21 Changeset [8500] by
-
#1861 applied persistence patches
- 09:45 Changeset [8499] by
-
#1886 improved crossover analyzer and added duplicate parent counting
- 00:15 Changeset [8498] by
-
#1886 added helper classes for generating charts
08/17/12:
- 14:32 Changeset [8497] by
-
Fixed behavior of VRP analyzers related to constraint relaxation (#1177)
- 11:54 Changeset [8496] by
-
#1886 removed deprecated code and cleaned up plugin dependencies
08/16/12:
- 16:48 Changeset [8495] by
-
#1902 additional change to 8494
- 16:33 Changeset [8494] by
-
#1902 changed mean and covariance function parameters to simple value parameters instead of constrained value parameters and changed the solution creator so that it doesn't throw an exception if no model could be found.
- 14:07 Ticket #1918 (Overflow in RunCollectionBoxPlot view) created by
- There is an unhandled exception in the box plot view. See the …
- 13:22 Ticket #1917 (In the CrossValidationView the value for the samples end field is not ...) created by
- Samples Start and Samples End remain the same even though a new CSV …
- 12:57 Changeset [8493] by
-
#1910: fixed handling of problem size changes
08/14/12:
- 17:35 Changeset [8492] by
-
- added TimeframeFeatureCorrelationView
- 17:10 Changeset [8491] by
-
#1902 fixed test cases, improved performance
- 15:26 Changeset [8490] by
-
#1081: Updated TimeSeriesPrognosisSolutionView.
- 15:24 Changeset [8489] by
-
#1902 removed caches in product and sum covariance functions and fixed bug in SE-ARD covariance function
- 14:31 Changeset [8488] by
-
- introduced weight property in Edge
- new data source implementation for DIMACS graphs
- lightweight data source implementation for osm graphs
- cost calculator interface and implementations added
- 14:31 Changeset [8487] by
-
#1081: Added ErrorCharacteristics View for time series models.
- 13:59 Changeset [8486] by
-
#1081: Corrected evaluators and time series models.
- 13:27 Changeset [8485] by
-
#1890 fixed a bug in the regression solution line chart that leads to an exception when the chart is shown which is however lost in the message queue. The bug has the effect that the empty point manipulation does not work correctly for instance in solutions produced through cross-validation
- 13:25 Changeset [8484] by
-
#1902 changed interface for covariance functions to improve readability, fixed several bugs in the covariance functions and in the line chart for Gaussian process models.
- 13:05 Changeset [8483] by
-
- Renamed ExtendedHeatMap to FeatureCorrelation
- deleted old CorrelationHeatMapView
- added FeatureCorrelationView
- 10:06 Changeset [8482] by
-
#1899 removed setting the checked state of certain analyzers in the samples unit test as this is already done correctly with the EnabledByDefault property of the analyzers
08/13/12:
- 17:37 Changeset [8481] by
-
- adapted AStar and Dijkstra algorithms for new graph representation
- test program modified
- 17:35 Changeset [8480] by
-
- fixed problem with edge category in XmlDataSource
- initial version for new DIMACS data source
- 17:31 Changeset [8479] by
-
- temporarily added weight and heuristic function to graph
- store category information and max speed with edge
- adapted Distance method
- 16:51 Changeset [8478] by
-
#1899: added population diversity analyzer which uses problem specific similarity calculators
- 16:18 Changeset [8477] by
-
- Added autoregressive target variable Symbol
- Merged trunk changes into the branch.
- 16:03 Changeset [8476] by
-
#1815: Corrected event registration in SymbolicExpressionGrammar and updated the variable view.
08/11/12:
- 14:45 Changeset [8475] by
-
#1902 fixed bug in calculation of variance in GPR model
- 10:34 Changeset [8474] by
-
- Set name of batch run to "repetitions times optimizer name"
- Added parent experiment for each instance when there are more than one instance and more than one variation per instance
08/10/12:
- 20:42 Changeset [8473] by
-
#1902 worked on GPR: added line chart, made parameters of mean and covariance functions readable, removed target variable scaling, moved noise hyperparameter for likelihood function to the end of the parameter list, added methods to calculate the predicted variance, removed limits for scale of covariance functions and introduced exception handling to catch non-spd or singular cov matrixes, implemented rational quadratic covariance function, added unit test case from GBML book (however it does not work as the book seemingly uses a noise-less likelihood function)
- 17:36 Changeset [8472] by
-
#1913: removed branch
- 17:35 Changeset [8471] by
-
#1913: integrated branch into trunk
- 17:07 Changeset [8470] by
-
#1913: increased point size and converted K to a fixed value parameter
- 16:39 Changeset [8469] by
-
#1916: prevented validation when view is readonly
- 16:27 Ticket #1916 (StringConvertibleValueView doesn't recognize it is read-only when ...) created by
- It will prevent the user from leaving the control forever if the …
- 15:58 Changeset [8468] by
-
#1081: Adapated calculation of results for horizons larger than 1.
- 15:38 Changeset [8467] by
-
#1913: fixed name of model
- 15:21 Changeset [8466] by
-
- Refactored classes
- Added parameter for number of iterations
- Added parameter for neighborhood sampling (allows to speed up gradient calculation)
- Adapted to changed k-NN algorithm
- 14:57 Changeset [8465] by
-
#1913: Changed k-NN to move model representation (kdTree) into the model object
08/09/12:
- 17:02 Changeset [8464] by
-
#1902 added const and noise covariance functions.
- 16:32 Changeset [8463] by
-
#1902 improved GPR implementation
- 16:20 Changeset [8462] by
-
- calculate distance in kilometers for two locations
- generate IGraph from a datasource
- adapted test program
- 16:17 Changeset [8461] by
-
- Implemented interface IGraph in Graph
- Equals method in vertex modified
- Equals method in Edge implemented
- New algorithm version for IGraph
- 15:01 Changeset [8460] by
-
#1081: Added horizon parameters to timeseries problem data.
- 14:36 Changeset [8459] by
-
#1081: Code clean up in time series prognosis model and problem data.
- 13:48 Changeset [8458] by
-
#1081: Derived time series classes from regression classes to avoid code duplication.
- 11:14 Changeset [8457] by
-
#1861 applied patches 0021 and 0022
- 10:56 Changeset [8456] by
-
#1861 fixed Pre and PostBuild events to work on Linux
- 09:33 Changeset [8455] by
-
#1902 changed calculation of gradients for covariance functions to reduce allocations of arrays
- 02:17 Changeset [8454] by
-
- Refactored NCAModel and NeighborhoodComponentsAnalysis algorithm
- Model now includes NearestNeighborModel
- Algorithm has ability to be canceled (basically recreated the optimization loop of mincgoptimize)
- Scaling should work properly now
- 01:44 Changeset [8453] by
-
#1481: Fixed plugin dependencies
- 01:31 Changeset [8452] by
-
#1905: Updated plugin dependencies
08/08/12:
- 23:59 Changeset [8451] by
-
#1861 merged changes from trunk into branch
- 23:34 Changeset [8450] by
-
#1861 some more renames and made post build event plattform-dependent
- 23:23 Changeset [8449] by
-
#1861 fixed mono compiler confusion with HL Random and System.Random
- 23:13 Changeset [8448] by
-
#1905: Corrected project references to ALGLIB 3.6.0
- 23:07 Changeset [8447] by
-
#1861 removed external evaluation from HL solution
- 22:52 Changeset [8446] by
-
#1861 fixed some more filenames and applied patch 0017
- 22:37 Changeset [8445] by
-
#1861 applied patch 0018
- 22:24 Changeset [8444] by
-
#1861 removed ProtocolBuffers from ExtLibs solution because the Mono compiler crashes when compiling it
- 22:13 Changeset [8443] by
-
#1861 applied patch 0016
- 22:04 Changeset [8442] by
-
#1861 applied patches 0014 and 0015
- 18:20 Changeset [8441] by
-
#1913: added quality output
- 18:15 Changeset [8440] by
-
#1861 made prebuild events os dependent
- 17:52 Changeset [8439] by
-
#1902 implemented unit tests for mean and covariance functions. implemented sum and product mean function. fixed incorrect gradient calculation in covprod
- 16:57 Changeset [8438] by
-
#1894 graph interface and implementation initial commit
- 16:46 Changeset [8437] by
-
- Improved speed of NCA
- Reorganized things
- 16:04 Changeset [8436] by
-
#1081: Merged changes from trunk into timeseries branch and refactored GP interpreter.
- 15:47 Changeset [8435] by
-
#1481: Added clustering solution evaluation view that performs PCA on the data to reduce it to two dimensions
- 14:12 Changeset [8434] by
-
- renamed old Vertex<T> to OsmVertex<T>
- added new Vertex class
- 14:12 Changeset [8433] by
-
#1892: Removed IConfigurableView from ScatterPlotView to avoid display of configure button
- 14:12 Changeset [8432] by
-
#1081: Corrected path of previous move operation.
- 14:08 Changeset [8431] by
-
#1081: Moved interpreter specific classes to separate directory.
- 14:08 Ticket #1915 (ClassNames caching leads to problems when dataset is not cloned) created by
- When a DataAnalysisSolution is created with the same ProblemData …
- 14:04 Changeset [8430] by
-
#1081: Intermediate commit of trunk updates - interpreter changes must be redone.
- 13:40 Changeset [8429] by
-
- renamed Graph to OsmGraph
- generic type in edge interface
- 13:16 Changeset [8428] by
-
#1081: temporary removed files to prepare for trunk update
08/07/12:
- 16:49 Changeset [8427] by
-
- Updated view
- Cloned ProblemData for solution
- 15:35 Changeset [8426] by
-
#1894 various error fixed
- 15:14 Changeset [8425] by
-
#1913: Added several initialization methods (LDA, PCA, and Random)
- 12:13 Changeset [8424] by
-
#1913: corrected tests and fixed bugs
- 10:52 Changeset [8423] by
-
- bidirectional version of Dijkstra algorithm
- method to get neighbors of a node in reversed order
- check for roundabouts in OneWay property
- 10:40 Changeset [8422] by
-
- Reduced memory footprint
- Fixed some bugs
- Added some tests
- 09:52 Changeset [8421] by
-
#1902 fixed default mean and covariance function for GP
- 00:42 Changeset [8420] by
-
- Worked on NCA
- Added scatter plot view for the model to show training data when it is reduced to two dimensions
It works, but I don't think it works correctly yet. I have randomized the initial matrix, because the starting point influences the achievable quality quite a bit.
