Machine Learning and The Big Data Revolution
3 likes3,670 views
The document discusses the integration of machine learning and big data within various industries, emphasizing how digitization can lead to unexpected network effects and growth. It outlines a repeatable methodology for applying data science to uncover new business opportunities through structured phases like data assessment and prototyping. The focus is on evolving data insights and improving business processes by leveraging machine learning and artificial intelligence.
Machine Learning and The Big Data Revolution
- 1. © 2016 IBM Corporation Machine Learning and The Big Data Revolution Rob Thomas VP, Product Development, IBM Analytics @robdthomas
- 2. © 2016 IBM Corporation2
- 3. © 2016 IBM Corporation3 Explicit Data…500 million individual data points A fertilizer and seed firm spends spends $1 billion on an IT company Data on nearly 5 million commercial properties
- 4. © 2016 IBM Corporation4 The History of Competitive Advantage Scale Effects 1 2
- 5. © 2016 IBM Corporation5 3 Superior insight from an evolving corpus of Data
- 6. © 2016 IBM Corporation6 Economies Of Scale Network Effects Data
- 7. © 2016 IBM Corporation7
- 8. © 2016 IBM Corporation8 Expand channels to engage users New economic models Individuals, not marketsRetail New sources of data Data applications New business processesFarming Data collection for advantage Redefine a skilled worker Social networks drive dataMedicine
- 9. © 2016 IBM Corporation9 How many industries are left, where if you digitize (or data-fy) some existing component of it, you introduce network effects in a way that was unanticipated, and leads to the potential for exponential growth? - Adaptation of Ben Thompson quote
- 10. © 2016 IBM Corporation10 Economies Of Scale Network Effects Data
- 11. © 2016 IBM Corporation11 Arthur Samuel, 1959 “Field of study that gives computers the ability to learn without being explicitly programmed.”
- 12. © 2016 IBM Corporation12 010101010101010101010101010101010101010101010101010101010101010101010 0101010101010101010101010101010101010101010101010101010101010101010101
- 13. © 2016 IBM Corporation13 Instead of programs that solve the problem, we need programs that learn to solve the problem.
- 14. © 2016 IBM Corporation14 Customer A Customer B Customer C Customer D DEPTH (but finite) BREADTH and infinite
- 15. © 2016 IBM Corporation15 Compute 100% Data Algorithms Compute Data Algorithms % of Budget The Last Decade Now
- 16. © 2016 IBM Corporation16 ML is designed for making predictions Making… Forecasting... Anticipating... Rating.. ….but, is this artificial intelligence?
- 17. © 2016 IBM Corporation17 Plan Machine Learning Deep Learning Artificial Intelligence Communicate Reason Take Action Learn
- 18. © 2016 IBM Corporation18
- 19. © 2016 IBM Corporation19 Make data simple Insights to out-think tomorrow Educating a new workforce
- 20. © 2016 IBM Corporation20 There is a repeatable methodology for applying the value of data science and machine learning in the context of your business, using your data to uncover new business opportunities. Phase 1 The Data Assessment Phase 2 The Workshop Phase 3 Application Prototype Understand data assets Understand data storage model, technology stack, tools and data science capabilities “Art of the possible” in data science and engineering Applying data science on your data asset backdrop Prototype an application to exploit new data science capabilities Reap benefits and re-invest in next project AnalyticsFirst
- 21. © 2016 IBM Corporation21
- 22. © 2016 IBM Corporation22 @robdthomas www.robdthomas.com http://tinyurl.com/bdrbook Rob Thomas VP, Product Development, IBM Analytics