Adaptive Windowing for Multi-Armed Bandits #71
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
shaharbar1
force-pushed
the
feature/adaptive_windowing
branch
9 times, most recently
from
035259b to
bdba9e4
Compare
shaharbar1
force-pushed
the
feature/adaptive_windowing
branch
6 times, most recently
from
0f98552 to
913968d
Compare
### Changes: * Added adaptive windowing mechanism to detect and handle concept drift in MAB models. * Introduced ActionsManager class to handle action memory and updates with configurable window sizes. * Refactored Model class hierarchy to support model resetting and memory management. * Added support for infinite and fixed-size windows with change detection via delta parameter. * Enhanced test coverage for adaptive windowing functionality across MAB variants.
shaharbar1
force-pushed
the
feature/adaptive_windowing
branch
from
913968d to
92df9b1
Compare
j3rom3c
reviewed
Feb 7, 2025
| window_length = len(rewards_window) | ||
| if window_length < 2: | ||
| return _NO_CHANGE_POINT | ||
| cumulative_reward = np.cumsum(np.array(rewards_window), axis=0) |
There was a problem hiding this comment.
I may be wrong but reward may be encoded -1/1 instead of 0/1 (in Hive I think). Did you make sure that the encoding was correct for cumsum?
j3rom3c
reviewed
Feb 7, 2025
| NonNegativeInt | ||
| The last change point for the given action. -1 if no change point is found. | ||
| """ | ||
| action_index = np.where([a == action_id for a in self.actions_memory])[0].tolist() |
j3rom3c
reviewed
Feb 7, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Changes: