DutchMLSchool. Machine Learning for Managers
0 likes181 views
This document outlines the themes and content of a workshop titled 'The Art of AI - Demystified,' co-organized by Jan W. Veldsink, covering areas such as machine learning, business model redesign, and the perceived benefits and challenges of AI implementation. It discusses the evolution of machine learning, decision engineering, and various applications in organizational settings, particularly in fields like fraud detection and customer churn prediction. The workshop also highlights the importance of governance, risk management, and the role of data science in informed decision-making.
DutchMLSchool. Machine Learning for Managers
- 1. Jan W. Veldsink MSc
- 2. Co-organized by: Sponsored by: Business Partners:
- 3. Jan W. Veldsink MSc 1rd edition | July 8 - 11, 2019
- 4. Jan W. Veldsink MSc THE ART OF AI AI - DEMYSTIFIED Jan W Veldsink MScâ¨
- 5. Jan W. Veldsink MSc
- 6. Jan W. Veldsink MSc Guidelines from the DNB
- 7. Jan W. Veldsink MSc AXVECO Technology Business model redesign Governance, risk & compliance Competences for sustainable innovation ⢠Business ecosystem ⢠Value chain analysis ⢠Revenue model ⢠Customer experience ⢠Process design and improvement ⢠Business rules/logic ⢠Systems architecture with new technologies ⢠Big Data Analysis, Robotics and AI ⢠Blockchain ⢠Smart contracts and dApp development ⢠GUI, legacy systems, infrastructure ⢠Operational risk, including privacy, cyber-security, business continuity ⢠Regulation and compliance ⢠Legality ⢠Governance implications ⢠Culture ⢠Leadership ⢠Capacity to innovate ⢠(Corporate) venturing ⢠Product development ⢠Agile, Lean Start up, design thinking ⢠Business case development
- 8. Jan W. Veldsink MSc Perceived benefits from AI
- 9. Jan W. Veldsink MSc Perceived benefits from AI
- 10. Jan W. Veldsink MSc âTHOSE WHO RULE DATA WILL RULE THE WORLDâ
- 11. Jan W. Veldsink MSc CHALLENGES OF AI
- 12. Jan W. Veldsink MSc Machine Learning Machine Learning is the steam engine of the XXI century When applied methodically, it helps solve complex problems at human-level performance When applied systematically, it can dramatically improve the performance of an organization (of all sizes and in all sectors).
- 13. Jan W. Veldsink MSc Machine Learning Evolution Genesis Custom built Product Service Utility Academics & Researchers Scientists Developers Analysts Everyone 1950s 1980s 2000s 2012 2030 Commodity 2020 Ubiquity CertaintyUnknown Defined NovelCommon Weka, R, Scikit, H2O BigML, Azure ML, Amazon ML, Google Cloud ML 1st Workshop on Machine Learning
- 14. Jan W. Veldsink MSc Predictive Services ⢠Every initiative is going to involve AI ⢠Every initiative is driven by DATA ⢠Every initiative is MODEL BASED â˘Every initiative is realised NON CODING
- 15. Jan W. Veldsink MSc Machine Learning Inputs Program Outputs Traditional Programming Inputs Program Outputs Machine Learning Itâs a CAT Label: Cat, dog, horse, spider Itâs another CAT
- 16. Jan W. Veldsink MSc Source of inspiration
- 17. Jan W. Veldsink MSc
- 18. Jan W. Veldsink MSc Source of Inspiration SYM BOLIC AI CONNECTIONIST AI Senses Classification Recognition Symbols Concepts Language Reasoning Planning
- 19. Jan W. Veldsink MSc Machine learned models
- 20. Jan W. Veldsink MSc Where to Start? Step 1 Finish âLetâs predict ⨠the Oscars!â âHere are the ⨠predicted winnersâ Step 2 - - - - - - - - ???
- 21. Jan W. Veldsink MSc Where to Start? Step 1 Finish âLetâs detect ⨠fraud! âHere are the transactions we should stop immediately. Step 2 - - - - - - - - ???
- 22. Jan W. Veldsink MSc Where to Start? Step 1 Finish âLetâs predict ⨠customer churn!â âHere are the customers we predict will leave our serviceâ Step 2 - - - - - - - - ???
- 23. Jan W. Veldsink MSc Where to Start? Step 1 Finish âLetâs predict ⨠diabetes!â âDoes a person have an indication for diabetesâ Step 2 - - - - - - - - ???
- 24. Jan W. Veldsink MSc
- 25. Jan W. Veldsink MSc BIGML meets Alexa & Lynx âWhat is the BMI?â âThe possibility that you become obese is 21,7%â â19â
- 26. Jan W. Veldsink MSc Decision tree â Diabetes classification
- 27. Jan W. Veldsink MSc TRENDS AND OBSERVATIONS
- 28. Jan W. Veldsink MSc ML Results on fraud
- 29. Jan W. Veldsink MSc Decision engineering
- 30. Jan W. Veldsink MSc Decision engineering ⢠Decision intelligence is an engineering discipline that augments data science with theory from social science, decision theory, and managerial science. ⢠Its application provides a framework for best practices in organizational decision-making and processes for applying machine learning at scale. https://en.wikipedia.org/wiki/Decision_Intelligence
- 31. Jan W. Veldsink MSc Decision engineering ML/AI expert Domain Data expert Business Domain expert
- 32. Jan W. Veldsink MSc Some Rabobank Projects DataDrift Agri Default Prediction Fraud CDD - Anomalies
- 33. Jan W. Veldsink MSc For example Current projects Some statistics: Number of users 85 Most non IT staff : â˘Fraud â˘AML â˘CDD â˘KYC â˘Credit Risk â˘Compliance auditing â˘Data integrity / Datadrift â˘Sanctions â˘Real estate dynamic pricing â˘Privacy violations
- 34. Jan W. Veldsink MSc Core Quesbons Viable How realistic is applying AI in this situation? (Is there a AI technology available for the current question) Valuable Does applying AI add value to customers / users? Vital How crucial is AI for achieving the result? AI / ML technology assessment
- 35. Jan W. Veldsink MSc
- 36. Jan W. Veldsink MSc THE ART OF AI AI - DEMYSTIFIED JAN VELDSINK MSC JAN@GRIO.NL J.VELDSINK@NYENRODE.NL
- 37. Jan W. Veldsink MSc