|Date||Time Start||Time End||Session Title||Speakers|
|5:00 p.m||5:30 p.m.||Comprehensive Data Labeling Technique (CDALT). JOIN ON 09/01 OR 09/02 |
Abstract: In order to take advantage of the power of machine learning methods, a system requires a large amount of data that is appropriately labeled in order to train a model. When data sets grow in size and complexity and the number of labels also grows, accessing data that is already labeled can be challenging and too expensive. To overcome this challenge and label a large dataset to train a machine learning model, we proposed “CDALT”, a semi-supervised and active learning framework which allows labelers to label their datasets fast and reliably.
|Federated Predictive Maintenance JOIN ON 09/01 OR 09/02|
Abstract: This demo showcases the benefits of a federated training for a predictive maintenance use case. Based on sensor measurement data, we are comparing four strategies for engine maintenance: corrective, preventive, predictive and federated. The demo aims to demonstrate the gain in predictive performance provided by the federated strategy compared to the three other strategies.
|Richard Vidal and Laura Degioanni|
|10:15 a.m.||10:35 a.m.||Computational Creativity For New Product Development JOIN ON 09/02 OR 09/03|
Abstract: The future is humans and machines working productively together, where machines don’t just automate but collaborate with humans in an impactful way. This demo helps us imagine being able to use AI like a team-mate, that helps in your product development process.
|10:30 a.m.||11:00 a.m.||Explainable AI for Banking: Counterfactual Explanations for Credit Applications JOIN ON 09/02 OR 09/03 |
Abstract: In recent years models have become deeper and more complex with resulting gains in predictive power. However this can come with a cost to transparency, the model is returning solutions but we don’t why. Explainable AI (XAI) aims at providing humans an explanation or intuition for these solutions. This demo will look at one such method called “Counterfactual Explanations”. By looking at small perturbations to the input vector the Counterfactual Explanations generate multiple “what if” scenarios that help explain why a certain automated decision was reached and what it would take to change that decision.
|09/02/2020||11:00 a.m.||11:30 a.m.||Scriptless Conversations JOIN |
Abstract: What if we could create a chatbot that could answer users’ queries only using a knowledge graph? This essentially means a paradigm shift from explicit conversational modelling techniques involving training of user utterances and templatizing of system responses. Introducing “Scriptless conversation” – Conversational AI using a knowledge graph. Our goal is to map free-form factoid, process-oriented and logical queries to a knowledge graph.
|4:30 p.m. |
|5:00 p.m. |
|AI for Life Sciences: Using Ampligraph to speed up the Drug Discovery Process JOIN ON 09/02 OR 09/03|
Abstract: This demo showcases how we modelled Genetic data in the form of a Knowledge Graph by using our in-house developed and open-source library, AmpliGraph, to learn the complex interactions and discover new associations between genes and diseases.
|5:15 p.m. |
|5:30 p.m. |
|AI for Workforce Reskilling: Using Knowledge Graph Embeddings to Guide Employee’s Choices JOIN ON 09/02 OR 09/03 |
Abstract: Shifting supply and demand for skills mean that mid-career reskilling of employees is an important concern for many companies and employees alike. We demonstrate the potential for Knowledge Graph Embeddings to capture contextual information about skills and work for the use case of inspiring and guiding employees selecting new-skilling options.
|09/04/2020||1:45 p.m.||5:00 p.m.||Measuring Algorithmic Fairness: challenges and solutions for the industry JOIN|| Laura Alvarez Jubete, Andreea-Roxana Miu, Medb Corcoran and Steven Tiell|
|09/04/2020||1:45 p.m.||5:00 p.m.||Knowledge Graph Embeddings: From Theory to Practice JOIN||Luca Costabello, Sumit Pai, Nicholas McCarthy and Adrianna Janik|
Enhancing Culinary Creativity with AI
Accenture Labs – Celebrating 30 Years of Innovation
Accenture – Tech4Good
Redefine the possible with applied R&D
Our 7 Accenture Labs are part of a continuous cycle where we identify and scale innovative solutions, ultimately bringing new opportunities to our clients. Focused on solving critical business problems with advanced technology, our 7 Labs brings fresh insights and innovations to our clients, helping them capitalize on dramatic changes in technology, business and society. We are a dedicated team of technologists and researchers who invest, incubate and deliver breakthrough ideas and solutions that help our clients create new sources of business advantage.
Results from Applied R&D can be experienced in seven R&D Hubs and dozens of Nano Labs across the globe. Our Labs locations are:
- San Francisco, CA
- Washington, D.C.
- Sophia Antipolis, France
- Dublin, Ireland
- Tel Aviv, Israel
- Shenzhen, China
- Bangalore, India
Our R&D Agenda:
- Artificial Intelligence
- Future Technologies
- Systems & Platforms
- Digital Experiences
- Application Engineering
- Cyber Security
Find out more: https://labs.accenture.com/