Speaker: Tom Osborn Seminar Date: Tuesday November 21 12:00pm Brief abstract: Active learning is a branch of semi-supervised machine learning for domains where data is expensive and where precise understanding is critical. Traditionally, active learning builds a model of a domain with a trade-off between exploration and exploitation probing of a domain to generate labelled data usefully. Earlier applications
Speaker: Andrew Stead, Firemark Labs Seminar Date: Tuesday November 14 12:00pm Brief abstract: The growing gap between customer expectations and what corporates are able to deliver is being filled by start-ups. Unencumbered by legacy systems and shareholder demands, start-ups are free to seek opportunities to create value for and own the customer. IAG has built Firemark to respond
Speaker: Roger Kermode (UTS) Seminar Date: Tuesday September 5 12:00pm Brief abstract: Recent research by the AIIA shows that many organisations are flying on auto-pilot when it comes to effectively using data and analytics for strategic purposes. Most of the effort locally in AI and Machine Learning focuses on operational and tactical issues which presents both risks (for incumbent
Speaker: Nejla Ghaboosi (IR) Seminar Date: Tuesday August 29 12:00pm Brief abstract: A substantial fraction of unsupervised learning research is driven by generative modelling. Generative Adversarial Networks (GANs) are a recently introduced class of generative models. This presentation tries to address: (1) why GANs are worth of studying, (2) how GANs compare to other generative models, (3) the details
Speaker: Timur Behlul, AlphaBeta Seminar Date: Tuesday August 1 12:00pm Brief abstract: Macro-economic data such as GDP typically undergoes many revisions prior to the final release. These revisions go on for several quarters, and can be quite significant. Yet many practitioners attempting to forecast current quarter GDP (now-casting) use the latest vintage of data. This is problematic, as using the
Speaker: Stella Xu Seminar Date: Tuesday June 20 12:00pm Brief abstract: Many of us may have heard of high frequency trading, but most of us may not realize that a large proportion of the ads that we see online are bought at real time as well. Demand side platforms such as The Trade Desk use different data sources to