DDD for Machine Learning

Transcript

1. Dr. Larysa Visengeriyeva Machine Learning Use Cases with DDD and

   ML Design Canvas

2. What kind of problems?

3. What kind of problems?

4. ML Use Cases Dimensions

5. When?

6. How?

7. Knowledge Crunching Event Storming For Finding AI Opportunities

8. Event Storming + ML Canvas

9. None

10. Finding Pain Points: Candidates for ML-Use Cases? ( 0DLOb HQWKLHOWb

    NHLQHJ¾OWLJHb 'DWHL 0Z6Wb HUNDQQW *LQL %LQ¦UGDWHLHQb 3')MSJb SQJ b H[SRUWLHUW 0DLOJXQ .DWHJRULHb HLQWUDJHQ (0DLOPLWb %HOHJDQb *RULOODb JHVFKLFNW 2&5b (UNHQQXQJ I¾Ub %LQ¦UGDWHL 0HWD'DWHQb -621b JHOLHIHUW (QWK¦OW)HOGb :HUWb .RPEL QDWLRQHQ %HOHJb HUVWHOOWDXVb *LQL-621 %HOHJb %HQXW]HU LQb ]XJHZLHVHQ %HOHJΖPSRUWSHU0DLO %HOHJΖPSRUWSHU8SORDG .DWHJRULHb PXVVDXWRPDW HUNDQQWb ZHUGHQ ΖFKZLOOPLWb GHP6\VWHPb QLFKWb LQWHUDJLHUHQ )XQNWLRQLHUWb QLFKWEHLb DOOHQb .DWHJRULHQ ΖFKZLOOPLWb GHP6\VWHPb QLFKWb LQWHUDJLHUHQ

11. Using Domains For Prioritization

12. How to Structure ML Projects

13. Structuring ML Projects

14. Structuring ML Projects

15. Value Propositions

16. ML Task (Start with a heuristic)

17. Data Sources

18. Features

19. Offline Evaluation

20. Decisions

21. Making Predictions

22. Collecting Data

23. Building Models

24. Live Evaluation and Monitoring

25. None