Data Science in the Real World: Making a Difference презентация
Содержание
- 2. Outline Making sense of World’s Data Building Data Systems Changing Dynamics
- 3. Michael Stonebraker “But then, out of nowhere, some marketing guys started
- 4. Michael Stonebraker “But then, out of nowhere, some marketing guys started
- 5. A Day in Your Life Think about a day in your
- 7. What can We do with Data? Optimize (World is inefficient) 30%
- 8. Building Data Processing Systems
- 9. Data Science Architecture
- 10. Data Processing Technologies Landscape
- 11. Batch Processing Store and process Slow (> 5 minutes for
- 12. Usecase: Big Data for development Done using CDR data People density
- 13. Value of some Insights degrade Fast! For some usecases ( e.g.
- 14. Complex Event Processing
- 15. Predictive Analytics If we know how to solve a problem, that
- 16. Usecase: Predictive Maintenance Idea is to fix the problem before it
- 17. Communicate: Dashboards Idea is to given the “Overall idea” in a
- 18. Communicate: Alerts and Triggers Detecting conditions can be done via Event
- 19. Case Study: Realtime Soccer Analysis
- 20. Changing Dynamics
- 21. Large Observational Datasets Stats are easy with designed experiments You
- 22. “It is better to be roughly right than precisely wrong.”
- 23. Challenges: Causality Correlation does not imply Causality!! ( send a book
- 24. Curious Case of Missing Data
- 25. More Data Beat a Clever Algorithm Observed by large internet companies
- 26. Challenges: Feature Engineering In ML feature engineering is the key [1].
- 27. Challenges: Taking Decisions (Context)
- 28. Challenges: Updating Models Incorporate more data We get more data
- 29. Challenges: Lack of Labeled Data
- 30. Two Takeaways Do your data Processing as part of a Bigger
- 31. Questions?
- 32. Скачать презентацию
Слайды и текст этой презентации
Скачать презентацию на тему Data Science in the Real World: Making a Difference можно ниже: