Conference Agenda
8:00-8:55 Registration. Pastries, Coffee and Tea
8:55-9:00 Opening Remarks by Nikolaos Vasiloglou
9:00-9:45 KEYNOTE ADDRESS: Executive perceptions of Machine Learning
-
How executives from various industries define machine learning
-
Is machine learning / artificial intelligence critical and why / why not?
-
What are the best applications for machine learning?
-
What are the biggest limitations and barriers?
-
The outlook of business executives on the future of machine learning
Beverly Wright
Executive Director, Business Analytics Center
Georgia Tech Business School Analytics Center, USA
9:45-10:10 Case study: Audience targeting solutions powered by data fusion and Machine
Learning
-
The changing advertising landscape
-
New data sources and the impact of Data Fusion
-
New advertising product offerings
-
Creating scalable solutions using Machine Learning
-
Results and Innovations
Wes Chaar
Senior Vice President, Analytics, Data and Decision Sciences
Turner/TimeWarner, USA
10:10-10:35 Case study: Machine Learning in mobile-phone based credit scoring at IBM
-
Insights from the largest PhD store on the planet – IBM
-
Boosted decision tree – a standard tool in any data scientist's toolkit
-
Using boosted decision trees to create credit scores for un(der)-banked individuals based on their recent cell phone usage
-
Demonstrating a lower default rate
-
Recommending the credit product to an additional 1 million customers compared to the bank's original method
Skyler Speakman
Research Scientist, IBM Africa
IBM, Kenya
10:35-11:05 Coffee Break
11:05-11:30 Case study: Prescriptive Machine Learning to optimise manufacturing processes
-
Traditional control philosophies do not uncover complex relationships between steps in manufacturing processes
-
Unsupervised machine learning techniques can be used to discover these relationships in order to reduce scrap, improve yield and lower operating expense variation
-
Case Study of DataProphet's solution and its results
Frans Cronje
Founder DataProphet, SA
11:30-11:55 Case study: Data-Driven operational efficiency for the enterprise
-
Major inefficiency sources across the different industry verticals
-
Why ML/AI? and Why now, not tomorrow?
-
Major roadblocks facing ML/AI-driven automation
-
Successful attempts for disrupting end-to-end value chains such as FMCG & Healthcare using scalable machine learning platforms
Mohamed Aly
Founder and CEO
Seeloz, USA
11:55-12:20 Deep Learning in Computer Vision
-
An introduction to convolutional neural networks
-
Understanding what’s possible by reviewing state-of-the-art approaches to various computer vision problems
-
Real-world case studies
Alex Conway
Founder and CTO, Data Scientist
NumberBoost, SA
12:20-13:50 Hot lunch catered by the Hilton Hotel
13:50-14:30 Google Cloud & Machine Learning
-
How Google uses machine learning to optomise internal processes
-
Practical example of machine learning in action at Google
-
How different industries are working with Google and Google Cloud (TensorFlow, ML Engine, and APIs)
David Elliott
Global Product Lead, Enterprise Cloud
Google, USA
14:30-14:50 Towards greater brand relevance, consistency and authenticity in the age of ML & AI
-
Brand tools: Finding nuggets of "predictive" knowledge in the waves of structured and unstructured data
-
The marketing data eco-system: automated data visualisation, content analysis, ML techniques and impact on marketing execution and digital marketing toolsets
-
Brand Consistency and brand AI and ML Innovation: creating organisational brand ambassadors through developing AI and ML – the role of algorithms and tools (Social Network Analysis, Conjoint Analysis, etc)
Paula Sartini
CEO
BrandQuantum, SA
14:50-15:15 A conversation about chatbots
-
Discovering different forms and 'personalities' of chatbots
-
Not every chatbot is made the same! Diving into their underlying architectures and capabilities to understand what makes them 'tick'
-
How chatbots can be relevant to existing businesses and fields, including current production systems and their impact
-
Practical advice and resources for relevant and accessible technology
Chris Currin
PhD Candidate in Computational Neuroscience
University of Cape Town, SA
15:15-15:40 Artificial Intelligence (AI) and economics. Skynet in the market
-
The impact artificial intelligence on economic theories.
-
Examples on how AI is applied to diverse areas such as modelling of the stock market, credit scoring, HIV and interstate conflict are given.
-
Artificial intelligence ideas used in this talk include neural networks, particle swarm optimization, simulated annealing, fuzzy logic and genetic algorithms.
Tshilidzi Marwala
Professor and Vice Chancellor
University of Johannesburg, SA
15:40-16:05 Coffee Break and Book Give-aways
16:05-16:30 Using analytics in insurance
-
How predictive analytics is changing the way insurers do business
-
Navigating the change from a traditional actuarial office to a modern multi-disciplinary team
-
Dealing with black box models and interpretability
-
Open-source vs proprietary software
Marc van Heerden
Senior Specialist, Advanced Analytics
Bain and Company, SA
16:30-16:55 Combining Behavioural Science and the Internet of Things with ML to increase
efficiency
Incorporating insights from behavioural science
-
Recent advances in behavioural science provide greater understanding of psychological factors that determine human behaviour
-
Well-designed experiments can be used to segment customer base (by psychological motivation, rather than observed behaviour)
-
Supervised classification used to map available client information on customer segments beyond experiment participants (Example: health insurer identifies psychological factors that cause deviations from intended dietary behaviour, and provides personalised, cost-effective interventions to augment behaviour)
Incorporating data from IoT
-
IoT used to measure location of workers and equipment
-
Supervised ML used to obtain flexible model of value added by combining specific equipment and workers
-
Instead of merely measuring value of purchasing equipment / hiring workers, this allows insights into whether current workers and equipment are employed efficiently (Example: IoT data from hospital theatre equipment and location tracking of nurses used to predict most efficient combination of staffs and equipment to maximise beneficial health outcomes and patient satisfaction)
Richard Barry, Polymorph Systems
16:55-17:20 Machine Learning for combatting cybercrime
-
Challenges of cybersecurity in the era of artificial intelligence
-
Methods and algorithms for fighting cybercrime
-
Practical examples
Nick Vasiloglou
Machine Learning Expert
Symantec, USA
17:20-18:00 Panel Discussion: In-house or outsourcing AI/Analytics in a large corporation. The role
of start-ups and consulting firms in the AI transformation of large enterprises.
Acquiring talent versus training human capital
Moderated by: Theunis Jansen van Rensburg
Head of Commercial Analytics
Wonga, SA
