Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

TensorFlow London 18: Dr Alastair Moore, Towards the use of Graphical Models in business modelling and technology strategy

229 views

Published on

Abstract: Recent developments in understanding technology diffusion and business strategy lend themselves towards analysis as directed graphs. Alastair will briefly introduce a Wardley Map, a directed dependency graph situated in a metric space. I will highlight aspects of this representation that lend themselves to analysis using dynamic graphical models. I will discuss some preliminary thinking about modelling aspects of a business, specifically those that are dependent on machine learning systems.

Bio: Alastair is a UCL Computer Science PhD (Computer Vision) with broad experience of applied Machine Learning and Statistical Analysis in a variety of settings. His background includes stints with corporate research, internet startups and universities. He was on the founding team on spin-out Satalia.com (Data Science) and venture backed WeArePopUp.com (Real Estate contracting) and helped setup the IDEALondon innovation centre with UCL and Cisco Systems. His primary objective at Mishcon de Reya is to ensure the firm consistently uses data and machine learning techniques to support and improve its business process, and helps advise its clients on innovation and technology. Alastair continues to maintain an active teaching role in the UCL School of Management (Predictive systems) the Cambridge Judge Business School (Entrepreneurship) and Peking University, Beijing (Technology innovation). His research interests include technology strategy, smart contracting and computational law.

Published in: Technology
  • Be the first to comment

TensorFlow London 18: Dr Alastair Moore, Towards the use of Graphical Models in business modelling and technology strategy

  1. 1. Towards the use ofGraphical Models in business modelling and technologystrategy A P Moore @latticecut TENSORFLOW LONDON #TENSORFLOWLDN
  2. 2. 19TH CENTURY MANAGEMENT Right now, your company has 21st-century [Internet]-enabled business processes, mid-20th century management processes, all built atop 19th-century management principles. - Gary Hamel
  3. 3. Business Models for 21st Centurybusinesses A P Moore @latticecut TENSORFLOW LONDON #TENSORFLOWLDN
  4. 4. MOTIVATIONS — Management ideas / tools have not progressed significantly – usually poorly integrated into business process systems — Software 2.0 - https://medium.com/@karpathy/software-2-0-a64152b37c35 Designing for constant evolution — Continuous improvement/integration human and machine systems
  5. 5. PROGRAM SPACE Source: https://medium.com/@karpathy/software-2-0-a64152b37c35
  6. 6. DROP OUT SYSTEMS DESIGN Source: Neural Networks and Deep Learning, Aurelien Geron, 2018
  7. 7. TUNE IN AND DROP OUT Would a company perform better if its employees were told to toss a coin every morning to decide whether or not to go to work? Well, who knows; perhaps it would! The company would obviously be forced to adapt its organization; - Aurelien Geron, 2017
  8. 8. PROGRESS
  9. 9. BUSINESS MODELS
  10. 10. WARDLEY MAPPING
  11. 11. SITUATIONAL AWARENESS
  12. 12. BUSINESS DESIGN PATTERNS
  13. 13. SATALIA
  14. 14. YOUR FIRST MAP
  15. 15. USER CENTRICITY
  16. 16. DEPENDENCY GRAPH
  17. 17. METRIC SPACE
  18. 18. MEASURE/ESTIMATE EVOLUTION
  19. 19. MANAGEMENT TOOL
  20. 20. Operators of large Data centres Existing software license model with highly personalized service Existing channel model and value add resellers. Enable our customers (operators of data centres) to more efficiently consumer power in their data centers Existing system modified to be provided on a hosted license basis (to prevent conflict with existing channels) Development of internal cloud based skills Support existing sales Hardware & cloud vendors providing infrastructure for new service. Existing distribution channel for sales High quality sensor Software setup Relatively low cost of software New marketing capability and more digitally focused sales team Existing license model maintained for both cloud and on premise (hybrid model).
  21. 21. Operators of large Data centres Existing software license model with highly personalized service Existing channel model and value add resellers. Enable our customers (operators of data centres) to more efficiently consumer power in their data centres Existing system modified to be provided on a hosted license basis (to prevent conflict with existing channels) Development of internal cloud based skills Support existing sales Hardware & cloud vendors providing infrastructure for new service. Existing distribution channel for sales High quality sensor Software setup Relatively low cost of software New marketing capability and more digitally focused sales team Existing license model maintained for both cloud and on premise (hybrid model).
  22. 22. The Machine Learning Canvas (v0.4)​ ​Designed for: ​ ​ Designed by: ​ ​ Date: ​ ​ Iteration: ​ . Decisions How are predictions used to make decisions that provide the proposed value to the end-user? ML task Input, output to predict, type of problem. Value Propositions What are we trying to do for the end-user(s) of the predictive system? What objectives are we serving? Data Sources Which raw data sources can we use (internal and external)? Collecting Data How do we get new data to learn from (inputs and outputs)? Making Predictions When do we make predictions on new inputs? How long do we have to featurize a new input and make a prediction? Offline Evaluation Methods and metrics to evaluate the system before deployment. Features Input representations extracted from raw data sources. Building Models When do we create/update models with new training data? How long do we have to featurize training inputs and create a model? Live Evaluation and Monitoring Methods and metrics to evaluate the system after deployment, and to quantify value creation. machinelearningcanvas.com​ by Louis Dorard, Ph.D. ​Licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
  23. 23. LEARNING LOOPS
  24. 24. GRAPHICAL MODELS MANAGEMENT TOOLS FOR 21ST C?
  25. 25. LEGAL ANALYTICS SEQUENCE FORECASTING
  26. 26. ACKNOWLEDGEMENTS Dr Louis Dorard AdjunctTeaching Fellow, UCL School of Management (UCL CS PhD) l.dorard@ucl.ac.uk @louisdorard Dr Dave Chapman Deputy Director, UCL School of Management d.chapman@ucl.ac.uk NikhilVadgama Deputy Director, UCL Centre for Blockchain Technology nikhil.vadgama@ucl.ac.uk Simon Wardley Leading Edge Forum @swardley

×