Commits · 91872f43020cb274508a838e9fe81da5a3403e44 · Nikolai.Hartmann / KerasROOTClassification

May 03, 2018
- Starting to mess with output scores plotting · 91872f43
  Eric Schanet authored 6 years ago
  
  Unverified
  
  91872f43
- Merge branch 'master' of gitlab.physik.uni-muenchen.de:Nikolai.Hartmann/KerasROOTClassification · 1b3e3fc3
  Eric Schanet authored 6 years ago
  
  * 'master' of gitlab.physik.uni-muenchen.de:Nikolai.Hartmann/KerasROOTClassification: Adding EarlyStop
  Unverified
  
  1b3e3fc3
- Same color for test and validation loss function · 133b7c96
  Eric Schanet authored 6 years ago
  
  Unverified
  
  133b7c96
- Adding EarlyStop · 423273cd
  Thomas Weber authored 6 years ago
  
  Stop training when a monitored quantity has stopped improving.
  423273cd
Apr 30, 2018
- debug plot for event weights · b4048320
  Alwin.Wester authored 6 years ago
  
  b4048320
- load model weights directly after initialising model, otherwise weights are reinitialised · b99f8435
  Nikolai.Hartmann authored 6 years ago
  
  b99f8435
- Write AUC into plot legends · 1c7e2228
  Thomas Weber authored 6 years ago
  
  1c7e2228
- Adding a few functions to compare setups · cbdbdefc
  Nikolai.Hartmann authored 6 years ago
  
  cbdbdefc
- Making loading and storage of data consistent · 9e229cfd
  Nikolai.Hartmann authored 6 years ago
  
  9e229cfd
- getters and setters for more datasets · a3ed26cf
  Nikolai authored 6 years ago
  
  a3ed26cf
- dump options to json dict · fa38a43d
  Nikolai authored 6 years ago
  
  fa38a43d
Apr 29, 2018
- Make optimiser and its arguments configurable · 1cce79f1
  Nikolai authored 6 years ago
  
  1cce79f1
Apr 27, 2018
- History, training and storage improvement · 11ff3e13
  Nikolai.Hartmann authored 6 years ago
  
  * store history information to json * shuffle training data before fit * include sample_weight * separate load function to plot from previously trained setup * weights and history are still stored after keyboard interrupt
  11ff3e13
- scores managed as properties (saved and loaded from h5) · 12b84267
  Nikolai.Hartmann authored 6 years ago
  
  12b84267
- Write auc in ROC plot · e827fe9c
  Thomas Weber authored 6 years ago
  
  e827fe9c
- Fixing typo · e5eafe08
  Eric Schanet authored 6 years ago
  
  Unverified
  
  e5eafe08
- Fixing ROC AUC computation · 2e2344e7
  Eric Schanet authored 6 years ago
  
  Unverified
  
  2e2344e7
- Selection argument, RobustScaler and workaround for numpy bug · 365ba76d
  Nikolai.Hartmann authored 6 years ago
  
  365ba76d
- Fixing ROC curve plotting · ce8f5544
  Eric Schanet authored 6 years ago
  
  Unverified
  
  ce8f5544
- Added plot functions for losses and accuracy · 77dd63a5
  Thomas Weber authored 6 years ago
  
  77dd63a5
- Remove unnescessary call to scaler · 1bb1f5f8
  Nikolai authored 6 years ago
  
  1bb1f5f8
- make method naming consistent (snake case) · 72aeb7c1
  Nikolai authored 6 years ago
  
  72aeb7c1
- Don't refit StandardScaler to test data · 738da982
  Nikolai authored 6 years ago
  
  738da982
- Merge remote-tracking branch 'origin/master' · 99966d64
  Nikolai authored 6 years ago
  
  99966d64
- Transform and plot data before training · 35d84c65
  Nikolai authored 6 years ago
  
  35d84c65
Apr 26, 2018
- Added swp files to .gitignore · a7e74f64
  Thomas Weber authored 6 years ago
  
  a7e74f64
- Added plotROC function and scaler function call · 660b4beb
  Thomas Weber authored 6 years ago
  
  660b4beb
- Keras MLP model and its training · 34930398
  Nikolai.Hartmann authored 6 years ago
  
  34930398
- Loading and saving of StandardScaler · 64b90646
  Nikolai.Hartmann authored 6 years ago
  
  64b90646
- load data from hdf5 if already existent · 2d10bf67
  Nikolai.Hartmann authored 6 years ago
  
  2d10bf67
- initial commit · 9ca36850
  Nikolai.Hartmann authored 6 years ago
  
  9ca36850