Commits · master · Sadjad Anzabi Zadeh / Reinforcement-Learning

22 Jun, 2022 1 commit
- Multi action PPO · 0f0d7e27
  Sadjad Anzabi Zadeh authored Jun 22, 2022
  
  0f0d7e27
19 Jun, 2022 3 commits
- stopping criteria is checked only if in traning · 198e671f
  Sadjad Anzabi Zadeh authored Jun 19, 2022
  
  198e671f
- Used serializable metrics · ea3e8335
  Sadjad Anzabi Zadeh authored Jun 19, 2022
  
  ea3e8335
- `MetricSerializerMixin` for keras metrics · 12e8823e
  Sadjad Anzabi Zadeh authored Jun 19, 2022
  
  12e8823e
16 Jun, 2022 8 commits
- Fixing circular reference issues · ed8f3a8c
  Sadjad Anzabi Zadeh authored Jun 16, 2022
  
  ed8f3a8c
- Finally, it works! · a725988e
  Sadjad Anzabi Zadeh authored Jun 16, 2022
  
  a725988e
- Minor bug fixes · 5a662d15
  Sadjad Anzabi Zadeh authored Jun 16, 2022
  
  5a662d15
- `LearnerProtocol` instead of `Learner` in annot. · 556ee7eb
  Sadjad Anzabi Zadeh authored Jun 16, 2022
  
  556ee7eb
- Implemented `_update_metrics()` · b57c7dd6
  Sadjad Anzabi Zadeh authored Jun 16, 2022
  
  b57c7dd6
- A PPO with warfarin specific metrics · ad82f7a5
  Sadjad Anzabi Zadeh authored Jun 16, 2022
  
  ad82f7a5
- Moved `SummaryWriter` and `metrics` to `Stateful` · 27f95e19
  Sadjad Anzabi Zadeh authored Jun 16, 2022
  
  27f95e19
- Added `_iteration` to learner · fd0ff875
  Sadjad Anzabi Zadeh authored Jun 16, 2022
  
  fd0ff875
15 Jun, 2022 7 commits
- `aggregate` accepts `n` (# of obs to agg) · 2547e662
  Sadjad Anzabi Zadeh authored Jun 15, 2022
  
  2547e662
- metrics · 17532f8f
  Sadjad Anzabi Zadeh authored Jun 15, 2022
  
  17532f8f
- metrics and summarywriter in agents · aae182bc
  Sadjad Anzabi Zadeh authored Jun 15, 2022
  
  aae182bc
- cleanups · 84161148
  Sadjad Anzabi Zadeh authored Jun 15, 2022
  
  84161148
- Implemented `SummaryWriter` · 4121ee68
  Sadjad Anzabi Zadeh authored Jun 15, 2022
  
  4121ee68
- A simple stopping criteria (similar to TF) · 709f09a4
  Sadjad Anzabi Zadeh authored Jun 15, 2022
  
  709f09a4
- MetricProtocol and two warfarin metrics. · e4086e90
  Sadjad Anzabi Zadeh authored Jun 15, 2022
  
  e4086e90
11 Jun, 2022 4 commits
- Used `add_iter()` instead of `add()` · a4619857
  Sadjad Anzabi Zadeh authored Jun 11, 2022
  
  a4619857
- Made `ActionRank` output 1-based · d00a644a
  Sadjad Anzabi Zadeh authored Jun 11, 2022
  
  d00a644a
- Added `amplifying_factor` to `NormalizedS...` · 0ce9c03d
  Sadjad Anzabi Zadeh authored Jun 11, 2022
  
  0ce9c03d
- new `add_iter()` method + improvements · 45dabea1
  Sadjad Anzabi Zadeh authored Jun 11, 2022
  
  45dabea1
05 Jun, 2022 1 commit
- Implemented few improvements · d0d4db44
  Sadjad Anzabi Zadeh authored Jun 05, 2022
```
Based on <openreview.net/attachment?id=r1etN1rtPB&name=original_pdf>
```
  d0d4db44
19 May, 2022 1 commit
- `PatientWarfarinOversampled` · ce8e62f2
  Sadjad Anzabi Zadeh authored May 19, 2022
  
  ce8e62f2
17 May, 2022 1 commit
- Critical bug fix in capped dose defs · 5cfeba17
  Sadjad Anzabi Zadeh authored May 17, 2022
  
  5cfeba17
30 Apr, 2022 1 commit
- `sensitivity` in `pa...full_dosing` · 704aba55
  Sadjad Anzabi Zadeh authored Apr 30, 2022
  
  704aba55
24 Apr, 2022 3 commits
- Bug fix in `get_config` · d294e62a
  Sadjad Anzabi Zadeh authored Apr 24, 2022
  
  d294e62a
- More metrics · fdcf04e9
  Sadjad Anzabi Zadeh authored Apr 24, 2022
  
  fdcf04e9
- Minor bug fixes here and there · 3c4c5922
  Sadjad Anzabi Zadeh authored Apr 24, 2022
  
  3c4c5922
20 Apr, 2022 3 commits
- Fixed save issue · 584841fc
  Sadjad Anzabi Zadeh authored Apr 20, 2022
```
`SystemRandom` cannot be saved. Used `Random` instead.
```
  584841fc
- Fixed an issue with multiple stoppings · d5dff611
  Sadjad Anzabi Zadeh authored Apr 20, 2022
  
  d5dff611
- Fixed save/load issues · c62d4b52
  Sadjad Anzabi Zadeh authored Apr 20, 2022
```
Had to remove `training` argument from `call`
```
  c62d4b52
17 Apr, 2022 6 commits
- compatibility and other minor changes · 17479661
  Sadjad Anzabi Zadeh authored Apr 17, 2022
  
  17479661
- Improved reil random generator code · d0437c18
  Sadjad Anzabi Zadeh authored Apr 17, 2022
  
  d0437c18
- used tf.function to make the model saveable! · 59a25f95
  Sadjad Anzabi Zadeh authored Apr 17, 2022
  
  59a25f95
- Impl. of 2010 + old 07 for future reference · aba38afa
  Sadjad Anzabi Zadeh authored Apr 17, 2022
  
  aba38afa
- Generator instead of random_seed · ca193929
  Sadjad Anzabi Zadeh authored Apr 17, 2022
  
  ca193929
- Backward compatibility fixes · 0fe13931
  Sadjad Anzabi Zadeh authored Apr 17, 2022
  
  0fe13931
27 Mar, 2022 1 commit
- Added normalized distance reward · 5a3a3858
  Sadjad Anzabi Zadeh authored Mar 27, 2022
  
  5a3a3858