Abstract
The present disclosure relates to methods and systems for efficiently storing data based on reinforcement learning and actor-critic models. An agent, comprising a “chunking actor” and a “redundancy critic” can segment various types of computer files into “dynamic segments” and “redundant segments” and store those segments in a dynamic datastore and a redundant datastore respectively. In general terms, redundant segments can comprise segments of files that exist “redundantly” among multiple files stored by the system, such as header information. Dynamic data segments can comprise data segments that are unique or generally unique to their respective files. By identifying redundant data segments, systems according to embodiments can avoid redundantly storing these data segments, thereby reducing data storage requirements. Upon request by a user, systems according to embodiments can reconstruct files by identifying and retrieving their dynamic and redundant data segments and provide those files to the user.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Bandyopadhyay, Debdeep; Roy, Alok; and Shetty, Prajna, "AI-Enabled System for Efficient Data Storage", Technical Disclosure Commons, (June 17, 2025)
https://www.tdcommons.org/dpubs_series/8244