Machine learning for product development
•
1 like•232 views
The document discusses machine learning and what is needed to build machine learning products. It begins with an introduction to machine learning, defining it as a computer program that improves at tasks through experience with data. It emphasizes that machine learning involves data collection, data pipelines, and measuring model performance. The key takeaways are that building machine learning products requires planning, experimenting in an iterative process, and using various machine learning frameworks and libraries while continuing to learn through online courses and communities.
Report
Share
Machine learning for product development
- 1. Machine Learning for product development
- 2. Agenda ● About me… ● How does AirHelp work? ● What is Machine Learning? ● What do I need to know to build ML products? ● Takeaways!
- 3. About me... • +10 years in the Software Industry. • Automation Product Manager at AirHelp. • Ubiquitous Computing, Distributed Computing and Machine Learning. • Large scale Machine Learning on production.
- 4. How does AirHelp work? Check your compensation for free. AirHelp enforces your lawful claim with the airline. Receive up to €600 in compensation.
- 5. What is Machine Learning?
- 6. What is Machine Learning? “A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P, if its performance at tasks in T, as measured by P, improves with experience E.” [Mitchel, 1997]
- 7. What is Machine Learning? “Instead of writing code, you feed data to the generic algorithm and it builds its own logic based on the data.” [Adam Geitgey, 2014]https://medium.com/@ageitgey/machine-learning-is-fun-80ea3ec3c471
- 8. What is Machine Learning?
- 9. Calculus, Linear Algebra, Statistics, Programming... What do I need to know to build ML products?
- 10. What do I need to know to build ML products? cutting edge innovation, you should know “a lot” cutting edge products, you should know “less”
- 11. business, engineers, data scientists. What do I need to know to build ML products? This is what I want and it needs to have Machine Learning! It’s a hype!
- 12. business, engineers, data scientists. What do I need to know to build ML products? I will use Machine Learning, just need a framework! But, wait, Feature Vectors, uh?
- 13. business, engineers, data scientists. What do I need to know to build ML products? Data, plot and a prediction model in R! But, wait, deploy it in Production? Runs on my machine!
- 14. https://hackernoon.com/the-ai-hierarchy-of-needs-18f111fcc007 What do I need to know to build ML products?
- 15. data collection as a culture. What do I need to know to build ML products?
- 16. What do I need to know to build ML products?
- 17. data and model pipelines. What do I need to know to build ML products?
- 18. data cleaning. What do I need to know to build ML products?
- 19. What do I need to know to build ML products? https://support.google.com/recaptcha#6081880 http://scikit-learn.org/stable/modules/svm.html
- 20. generate data thinking on models. What do I need to know to build ML products?
- 21. old approaches don’t hurt. What do I need to know to build ML products?
- 22. new approaches are great. What do I need to know to build ML products?
- 23. http://knowyourmeme.com/photos/1244486-what-people-think-i-do-what-i-really-do What do I need to know to build ML products?
- 24. measure your models, always. What do I need to know to build ML products?
- 25. What do I need to know to build ML products?
- 26. focus on business goals. What do I need to know to build ML products?
- 27. What do I need to know to build ML products? https://hackernoon.com/the-ai-hierarchy-of-needs-18f111fcc007
- 28. success. What do I need to know to build ML products?
- 29. What do I need to know to build ML products?
- 30. experience E. features, labels... tasks T. supervised, unsupervised... performance P. f-score, AUC... Takeaways!
- 31. plan. do. act. check. Takeaways!
- 32. jupyter, keras, tensorflow, pandas, spark mllib, libsvm, opencv, gensim, pytorch, h2o, mxnet, numpy, and many more in your nearest store. Takeaways!