Ad exchange bid optimization with reinforcement learning
US11645679B2 · kind B2 · utility
1Cited by
0References
1Claims
0Family size
Assignee
Inventors
Key dates
| Filing date | Jul 19, 2021 |
| Grant date | May 9, 2023 |
| Priority date | — |
| Expiry date | Jul 19, 2041 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG06Q30/0276
- WIPO fieldIT methods for management
- WIPO sectorElectrical engineering
Abstract
A system for training a bidding model comprising: a plurality of tactics stored on at least one database; a plurality of hyperparameters; in response to an available inventory from a publisher relayed through a real time bid server, computing a bid on the available inventory; sending the bid to the real time bid server; receiving an auction result in response to the bid; calculating a plurality of rewards based on the auction result and the tactics; calculate a plurality of q values based on the rewards; calculate a plurality of losses; backpropogating the losses through the bidding model.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.