Slideshare uses cookies to improve functionality and performance, and to provide you with relevant advertising. If you continue browsing the site, you agree to the use of cookies on this website. See our User Agreement and Privacy Policy.

Slideshare uses cookies to improve functionality and performance, and to provide you with relevant advertising. If you continue browsing the site, you agree to the use of cookies on this website. See our Privacy Policy and User Agreement for details.

Successfully reported this slideshow.

Like this presentation? Why not share!

- 2012 Big Data - Bigger Problems for... by Ayasdi 972 views
- Allison Gilmore, Data Scientist, Ay... by MLconf 1437 views
- Why Topological Data Analysis Beats... by DataRefiner 5184 views
- Ayasdi strata by Alpine Data 1572 views
- Machine Learning with Ayasdi by Ayasdi 2723 views
- Topological Data Analysis: visual p... by DataRefiner 4584 views

1,245 views

Published on

Published in:
Technology

No Downloads

Total views

1,245

On SlideShare

0

From Embeds

0

Number of Embeds

8

Shares

0

Downloads

56

Comments

0

Likes

1

No embeds

No notes for slide

- 1. Shape as Organizing Principle for Data MLConf Seattle 2015 Anthony Bak, Principal Data Scientist
- 2. The Data Problem: Complexity
- 3. Solution: Topological Summaries
- 4. Shape as Organizing Principle for Data
- 5. Shape as Organizing Principle
- 6. Reduce Bias, Discover Models TDA tells you the data you have, not the data you want to have.
- 7. Generating Topological Summaries
- 8. Generating Topological Summaries
- 9. Generating Topological Summaries
- 10. Generating Topological Summaries
- 11. Generating Topological Summaries
- 12. Generating Topological Summaries
- 13. Generating Topological Summaries
- 14. Generating Topological Summaries
- 15. Generating Topological Summaries
- 16. Generating Topological Summaries
- 17. Generating Topological Summaries
- 18. Generating Topological Summaries
- 19. Generating Topological Summaries
- 20. Generating Topological Summaries
- 21. Generating Topological Summaries
- 22. Generating Topological Summaries
- 23. Generating Topological Summaries
- 24. Remember/Forget Use multiple lenses/metrics to get the complete picture Different lenses provide different summaries
- 25. Generating Topological Summaries
- 26. Lenses: where do they come from? Mean/Max/Min Variance n-Moment Density … Statistics PCA/SVD Autoencoders Isomap/MDS/TS NE … Machine Learning Centrality Curvature Harmonic Cycles … Geometry
- 27. Why Topology?
- 28. Key Properties of TDA Deformation Invariance Compressed Representation Coordinate Freeness
- 29. Coordinate Invariance 1. Topology of shape doesn’t depend on the coordinates used to describe the shape 1. Different feature sets can describe the same phenomena 1. While processing data, we frequently alter coordinates: scaling, rotating, whitening You want to study properties of your data that are invariant under coordinate changes
- 30. Coordinate Invariance: Gene Expression NKI GSE230
- 31. Coordinate Invariance: Disease State
- 32. Deformation Invariance • Topological features don’t change when you stretch and distort the data Advantage: Makes problems easier Noise resistance Less pre-processing of data Robust (stable) data
- 33. Deformation Invariance
- 34. Deformation Invariance
- 35. Deformation Invariance
- 36. Deformation Invariance
- 37. Compressed Representation • Replace the metric space with a combinatorial summary: a simplicial complex. • Data becomes easier to manage, search, and query while maintaining essential features. • Leverages many known algorithms from graph theory, computational topology, computational geometry.
- 38. Compressed Representation
- 39. Baby Steps: PCA
- 40. PCA
- 41. PCA
- 42. Data Stories
- 43. Model Introspection
- 44. Model Introspection
- 45. Predictive Maintenance
- 46. Customer Churn
- 47. Customer Churn
- 48. Transaction Fraud
- 49. Transaction Fraud
- 50. Transaction Fraud
- 51. We’re Hiring! http://www.ayasdi.com/company/careers/ Data Has Shape And Shape Has Meaning

No public clipboards found for this slide

Be the first to comment