I guess at first the data updates are more frequent until you get to a point where more data doesn’t give you better performance. At this point the frequency of updates decreases, but how do we know how often we should update the data then?
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Best way to benchmark your Data-centric AI? | 1 | 58 | May 16, 2023 | |
| Should data augmentation be encouraged? | 1 | 57 | May 18, 2023 | |
| Decision on data quality and quantity | 1 | 64 | August 11, 2021 | |
| Best principles for good Data Centric approach? | 1 | 59 | May 18, 2023 | |
| Refine or train from scratch after updating dataset | 1 | 43 | May 18, 2023 |