Actions and Detail Panel
Data Science Lab: Understanding Graphs & Tensor Flow
Wed 7 December 2016, 18:30 – 21:00 GMT
Welcome to our last Data Science Lab event of 2016!
This evening will be themed around graphs and graph databases, and understand how to use Tensor Flow.
18:30 Doors open
19:00 Talk: What can knowledge graphs do for me?
Speaker: Dr Sheldon Hall
As the world becomes more and more data rich the difficulties of managing and wrangling data become more pronounced. The knowledge graph is positioned to be the next thing big thing in terms of data integration and search, but what is it and where can you get one? The next question might be: should I care? The answer is a definite YES and in this presentation I will justify this claim by highlighting the benefits to data scientists. These include: easy integration of multiple data sources, a flexible schema to check correctness and express how the information is related, and an intuitive graph query language to extract information across all sources.
Sheldon Hall is a software engineer working at Grakn Labs where a distributed knowledge graph is being developed. His current role involves research and development of our graph based knowledge representation system and analytics engine. He has a master’s degree in mathematical modelling and scientific computing, a doctorate in nuclear engineering, and is a previous ASI Fellow.
19:45 Talk: The flow in TensorFlow
Speaker: Dr Alberto Favaro
Learning TensorFlow is much more than becoming familiar with different names for known commands. New users typically face the challenge of understanding the general ideas that motivate the very structure of TensorFlow. Indeed, this software library is characterised by strong design choices — such as the representation of algorithms via computational graphs — that require some getting used to.
In view of these considerations, this talk will introduce TensorFlow by describing its main philosophical principles. The benefits that result from accepting the new framework, in terms of distributed computing and model training with data batches, will be highlighted and discussed. To complement the theory with a basic example, participants will be guided through the implementation of a neural network in TensorFlow.
Alberto is a data scientist and theoretical physicist and a visiting scholar at Imperial College London. He recently joined the Consultancy Team at ASI Data Science, where he now works on the commercial applications of machine learning. His present activities include developing a biometric keystroke recognition system, implementing neural networks for image classification with Raspberry Pi, and writing TensorFlow demos for educational purposes.
20:15 Drinks & Networking
21:00 Event finishes