wise.io PyData SV 2013: Machine Learning from Fast to Easy

An overview of the wise.io machine-learning tools including the super fast WiseRF machine learning package (working locally, on EC2 or even on a Raspberry Pi) and the upcoming machine-learning service with automatic model optimization and 1-click addition of domain expertise.

Transcript

1. Henrik Brink, CTO PyData SV 2013, March 19 Machine Learning, Python and Raspberry Pi’s @brinkar @wiseio
4. 11kly 11kx Reference New Difference Palomar Transient Factory (PTF) ~1.5 million candidates per night ~10 new transients http://www.astro.caltech.edu/ptf/
5. PTF11kly (SN 2011fe) ©Peter Nugent Supernova Discovery in the Pinwheel Galaxy 11 hr after explosion Nearest SN Ia in >3 decades
6. http://bigmacc.info
8. Python Bootcamps at Berkeley
9. Machine Learning in
10. ML pain points I • Data is messy! • Hard to scale non-linear algorithms to large datasets • Ad-hoc feature engineering • Collaboration on data, features and models is difficult data scientists
12. Random Forest® Random Forest® is a registered trademark of Salford Systems
13. Brink+2012
14. Brink+2012 10% label noise
15. http://wise.io
17. bit.ly/YZZb9d WiseRF™
18. WiseRF™ 12 GB MNIST in 90 seconds 8 core EC2 instance
19. WiseRF™ A brain for the Internet of Things!
20. Demo™ WiseRF™
21. Higher speeds Larger datasets Hadoop / Mahout Energy sensors Ad bidding Credit card fraud detection Video tracking Financial predictions Batch product recommendations Real-time recommendations Internet of Things Healthcare sensors Fast vs Scalable
22. High-frequency data science High-frequency prediction Feature engineering Data ingestion Model deployment Model validation High-frequency Machine Learning
23. • Statistical validation of models • Lack of feature engineering expertise • Dealing with data and computing infrastructure ML pain points II application developers
24. Machine Learning as a Service™
25. Machine Learning as a Service™ bit.ly/15eWEZG
26. Reusable features
27. Collaboration
28. Integration
30. docs.wise.io
34. • Scalable infrastructure required • Hard to go from data science experiments to production. • Complete privacy / security. ML pain points III business and enterprise
37. @brinkar @wiseio henrik@wise.io We’re hiring! contact@wise.io