Coffee and Registration
Start of conference – Chairperson’s introduction and welcome
First Session Chairperson: Professor Gautam Mitra, OptiRisk Systems/UCL -
Keynote: Foundations of Deep Learning, illustrated with use cases
Bianca Furtuna, Data Scientist at Elastacloud
What is the future going to look like? When are we going to reach true Artificial Intelligence? Is the Singularity going to happen? There is a lot of talk today about AI and what it means for human society. Let's forget about the future for now and focus on what is possible today. We are going to look at the most promising area in AI research, Deep Learning and understand how it fits in the wider picture of Machine Learning. We are going to explore the fundamentals of Deep learning and deep dive into some common use cases to illustrate the applications of the technology in the real world.
Artificial Intelligence and Machine Vision: The next frontier for process efficiency and integrity
Bogdan Ciubotaru, CTO, Everseen Ltd
♦ Process efficiency and integrity represent two main factors impacting many sectors including retail, manufacturing and transportation
♦ Vision has always been a critical information source, however it is mainly specific to human observers and hence difficult to use effectively
♦ Machine vision and artificial intelligence open the door for a wide range of applications including efficient process management
♦ The positive impact on various businesses and industries has already been proven with great growth envisioned for the future
Second Session Chairperson: Jochen Leidner, Director, Research, Thomson Reuters -
Becoming Cognitive: How to Transform Your Business
Josh Sutton, Global Head, Data & AI, SapientRazorfish
AI is going to disrupt nearly every industry at a faster pace than we have ever seen. Tomorrow’s success stories will be those firms that became a cognitive business. This presentation will discuss pragmatic, real world approaches for identifying meaningful uses for AI within your organisation today. It will outline the seven steps for cognitive transformation within an enterprise business. The audience will gain a high level understanding of how to build a cognitive platform for their organisation inclusive of technology, experience, and change management that avoids creating silos and demonstrates meaningful business value in months instead of years.
AI in Production: How super-intelligence becomes dumb
Alistair Ferag, Senior Data Scientist, Satalia
Satalia creates production-grade data science and optimisation solutions for a range of clients and ever aspire to develop AI into them - but this can be tough. Leaving AI solutions in production with no oversight can lead to unintended consequences. This talk will discuss approaches to apply AI and provide a demo to highlight how difficult it can be to operationalize.
The most critical selection criterion
Gert De Geyter, Senior Consultant Data Analytics at Deloitte Belgium
As data scientists, we often spend a long time optimizing and endlessly trying to refine our models. All too often, this ends up in neglecting what may be the most important selection criterion: acceptance by the end user. In this talk, some tips and tricks are shown using real case examples how to improve to odds of convincing critical end users.
Comparing Notebooks for Data Science
Karlijn Willems, Data Science Journalist at Datacamp
Data science requires more than traditional Integrated Development Environments (IDEs) can offer: the need to create and share data stories. That's why data scientists often resort to notebooks. In her talk, Karlijn Willems will guide you through the landscape of data science notebooks, from Jupyter to Beaker to R Markdown to Zeppelin and more, providing a comparison between the different notebooks that are out there for data science enthusiasts!
Speaker Panel and Q&A Session
Third Session Chairperson: Jochen Leidner, Director, Research, Thomson Reuters -
Intelligent Information: R&D and Innovation in Information Access at Thomson Reuters
Jochen Leidner, Director, Research, Thomson Reuters
The Information economy combined with progress in computer performance and progress in machine learning pose great opportunities. In this talk, I will give some case studies of research projects conducted at Thomson Reuters Corporate Research & Development group, where we strive to improve information access for professional knowledge workers in different vertical domains, often applying machine learning to applications in information and information retrieval.
Using Data Science for Underwriting and Risk Analytics for Unsecured Instalment Loans
Aditya Satyadev, Co-founder & CEO, BizAcuity Solutions Pvt. Ltd.
♦ Unsecured Instalment Loan Business and Complexity
♦ Conventional rule based Underwriting Vs. Data Science Based Underwriting
♦ Paradigm shift in Risk Management due to accessibility of in-house and global data
♦ Solution Architecture – Data Integration/Wrangling, Data Quality, Technology Stack
♦ Building the Model with Data Science and Deep Learning Algorithms
♦ Model Assessment and Optimization
♦ Business Impact Analysis
♦ Operationalization of Solution with Human Touch
Deep Learning is back and how enterprises can leverage it!
Tarry Singh, Data Analytics Executive, Entrepreneur
"Tarry will give a whirlwind tour of the world of deep learning. How it all started -- yes, your linear algebra and spherical trigonometry is back. Explore the inner workings of how Deep Learning actually works -- how ANNs work and how they still have a long way to go before really understanding how human brain works. Then he will take a practical dive into how companies actually try to put this is practice and create great products and services. And if time permitting he will give a quick tech walkthrough into one of his AI projects from his upcoming book titled 'Practical AI / Deep Learning Projects' "
Speaker Panel and Q&A Session
Final Session Chairperson: Professor Gautam Mitra, OptiRisk Systems/UCL -
Networks are like onions: Practical Deep Learning with TensorFlow
Barbara Fusinska, Data Scientist
Deep learning is the area that wins over the field of Artificial Intelligence. By using libraries like TensorFlow, it is now available to the wider audience. In this tutorial, Barbara will walk the audience through the process of creating several types of neural networks. The session will start with explaining key concepts of deep learning and introducing datasets the computation will be performed on. Along the way, attendees will have the practical opportunity to use TensorFlow to build deep networks, train them and evaluate the results. After the session, participants will become familiar with how to use TensorFlow when shaping the architecture of neural networks. By the hands-on form of the tutorial, the audience will have the chance to gain some firsthand experience of how to apply deep learning to computer vision and natural language processing tasks.
Applications of Deep Learning in Business
Armando Vieira, Data Scientist, ContextVision AB
Despite being a relatively new research field, Artificial Intelligence (AI) history has been shaped by huge expectations and colossal failures. After several stagnation periods or "long winters", AI is flourishing as impacting business at an unforeseen pace. Behind this success is a technology widely known as Deep Neural Networks or Deep Learning (DL). In this talk Armando summarizes the key elements of DL and why it is such a transformative force for almost every business.