Dear Q Group members,

You are invited to attend a seminar “Machine Learning Applied in Finance: recovering latent connections between time-series data” given by Dr Amir Dezfouli (CSIRO Data61), at 5pm Thursday 14 March 2019, at Data61 Demonstration Lab (710 Collins Street, Docklands) in Melbourne.

Agenda: Thursday 14 March 2019

5:00pm Pre-drinks and food

5:30pm Seminar starts

6:30pm Networking

Abstract: Given a set of time-series, (e.g, property prices over time or stock indices over time) I will present a method for discovering connections (e.g., associations or causal interactions) between them. The method is presented within the area of network discovery in which each node presents a time-series and we are interested to find the latent arcs (connections) in the network. Traditional methods for the discovery of latent network structures are limited in two ways: they either assume that all the signal comes from the network (i.e. there is no source of signal outside the network) or they place constraints on the network parameters to ensure model or algorithmic stability. We address these limitations by proposing a model that incorporates a Gaussian process prior on a network-independent component and formally proving that we get algorithmic stability for free, while providing a novel perspective on model stability as well as robustness results and precise intervals for key inference parameters. I will also present the results of applying the method in different domains such as finance.

Bio

Amir earned his bachelor’s degree in software engineering and his master’s in artificial intelligence from the University of Tehran in 2006 and 2009 respectively. He then earned his PhD from the University of Sydney in 2015 and subsequently was a post-doctoral researcher at the University of New South Wales working on the analysis of high-dimensional time-series data. He is currently a research scientist at the machine learning research group, Data61, CSIRO, working on method development in various machine learning domains such as probabilistic inference and reinforcement-learning.

Registration:

Please use the following link for RSVP

mailto:qrsvp@qgroup.org.au?subject=[Melbourne]_I_would_like_to_attend_seminar_14_March_2019

Regards,

The Q Group Committee

_______________________________________________
You may update your details by visiting https://qgroup.org.au/user. (Your user id and password are required.)